After Pearl Harbor: The Proper Role of Population Data Systems in Time of War 

by William Seltzer (Fordham University) and Margo Anderson (University Wisconsin -- Milwaukee)

 

NOTE: The present draft was completed on March 8, 2000 to serve as the basis for our presentation at the session on "Human Rights, Population Statistics, and Demography: Threats and Opportunities," Population Association of America, Annual Meeting, March 23-25, 2000, Los Angeles, CA. An electronic copy of the paper was provided to the session chair and discussant and to the U.S. Census Bureau on that date with the notation on the cover sheet "Not for quotation, direct or indirect, prior to March 25, 2000, 8 am PST." Even though we circulated no other copies of this draft and despite the embargo, reports based on it began appearing in the press on March 17. Both at an advance presentation of the paper at the Census Bureau on March 15 and during its presentation at the Population Association meetings we received many helpful comments and criticisms. We are circulating the paper now to obtain comments and suggestions for improvement more broadly as well as to provide interested colleagues an opportunity to go beyond the partial reports of our work that appeared in the media. Information on contacting either author is presented on the next page. Subsequently, we will revise the paper to take into account the comments and criticisms received.

ABSTRACT

Following other studies of the misuse of otherwise benign population data systems to assist in the perpetration of major human rights abuses, we present interim results of our study of how such data systems were used in this country in the early stages of World War II. We trace the involvement of the federal population statistics system, including individual agencies, outputs, staff and advisory committees, in a series of war-related activities. These activities include the internment of over 100,000 Japanese Americans then living in the Pacific Coast States, the alien registration of 1940, the registration and tracking of "enemy aliens" early in the war, and an unsuccessful effort to establish a general population registration system for military and statistical purposes. We also briefly examine the possible long-term consequences of some of these activities and discuss the research and human rights and statistical policy implications of our findings.

CONTACT INFORMATION

William Seltzer
Department of Sociology and Anthropology
Dealy 407
Fordham University
441 East Fordham Road
Bronx, NY 10458

email: seltzer@fordham.edu

Margo Anderson
Department of History
University of Wisconsin - Milwaukee
Milwaukee, WI 53201

email: margo@uwm.edu


I. Introduction

Following other recent studies of the misuse of otherwise benign population data systems to assist in the perpetration of major human rights abuses [Aly and Roth, 1984; Remond et al, 1996; Seltzer, 1998, 1999a; Søbye, 1998], this paper provides interim results of our continuing examination of how such data systems were used to assist in the military defense of the United States in the early stages of World War II and the human rights and statistical policy implications of this involvement. A major aspect of this examination is the often-discussed issue of how the US Census Bureau contributed to the internment of some 100,000 Japanese Americans then living in the Pacific Coast States. Nevertheless, our concern is somewhat broader. After dealing with evidence and issues related to the internment program (section II), we briefly trace the development of registration systems for the military draft and for aliens and continue with an examination of the programs for the control and registration of enemy aliens on the West Coast early in 1942 (Section III). We next describe an ultimately unsuccessful effort to establish a general population registration system throughout the United States for military and statistical purposes (section IV). We also briefly examine the question of whether any long-term consequences of these events can be identified (section V).

After this largely factual account based primarily on our reconstruction of events that occurred nearly 60 years ago, we explore the apparent motivations of the demographers and statisticians associated with these events and a series of critical decision points (section VI). In addition, and equally important, in the same section, we present some implications of this historical review for statistical policy and research issues for today and the future.

Given that this paper is being presented within a week of the start of enumeration of the 22nd decennial census of the population, one conclusion we would stress at the outset is that nothing in this paper should be taken as discouraging the need for each person and family to be included in the forthcoming census count. Indeed, as one of us has previously emphasized [Seltzer, 1998: 539-540], a decennial population census is, next to the sample survey, generally the least threatening method of collecting needed population statistics. In this context, the recent efforts of the Census Bureau to increase the use of sampling in connection with the decennial census can only be viewed as highly salutary from the perspective of prudent human rights policies.

Finally, a personal note: Both authors of this paper feel a strong attachment to the Census Bureau as an important public institution performing a valuable social function, often under the most difficult circumstances. Regrettably, some of our findings and conclusions call into question certain Census Bureau actions and policies and those of others active in the federal statistical community. We are aware that hindsight can make us all wiser and that during the immediate post-Pearl Harbor period actions were done by many in the interests of patriotism and misguided self-defense that were ultimately recognized as wrong. However, to leave such mistakes unexamined bars us from learning from them and perhaps avoiding similar errors in the future. Although our research documents what we consider to have been both past and continuing mistakes, we have also identified a few instances where, even in the context of the crisis atmosphere of early 1942, Bureau staff were alert to the possible misuse of population data systems.

II. The Internment of the Japanese Americans

We are not the first persons to have reviewed the issue of the involvement of the US Census Bureau in the internment of the Japanese Americans during Word War II. The subject was dealt with first by the Bureau itself in a section of its unpublished War History volume [Dobbin, 1946a]. Subsequently, several investigators of the Japanese American internment experience have touched upon various aspects of the issue in the course of their research, sometimes locating highly relevant archival materials [see particularly, Okamura, 1981; Daniels, 1982; and Weglyn, 1996]. In addition, over the intervening decades, newspaper stories have appeared often based on or inspired by the results of these research efforts [San Francisco Examiner, 1981], raising the issue of the nature of the Bureau's involvement in the internment effort. Largely in response to the questions raised by these researchers and the related press coverage, the Bureau has provided a series of explanations of its involvement in the internment program, sometimes citing its own archival research [see, for example, Bohme, 1975; Barabba, 1981; Dedrick, 1981a and b; Chapman, 1982; US Census Bureau, 1982; and Bohme and Pemberton, 1991]. However, all prior examinations of this issue have been carried by persons who either (i) lacked knowledge of population statistics operations and the history of the Bureau, but who were independent of the Bureau or (ii) had such knowledge, but were not independent of the Bureau. It is hoped that our combination of expertise and independence, may help to provide some new insights into these events of nearly 60 years ago.

It maybe useful to begin by offering three worse-case hypotheses concerning the involvement of the US Census Bureau in the internment of Japanese Americans after the US entry into World War II:

(1) That the Census Bureau provided names and addresses of Japanese Americans obtained from the 1940 Population Census files, in violation of census confidentiality laws and related assurances, to the military authorities to assist in the internment program;

(2) That the Census Bureau has engaged in a deliberate and systematic cover-up of its involvement in the internment program; and

(3) That the very existence of the category, "Japanese," within the racial classification used by the Census Bureau in the 1940 Census, led to a different treatment of Japanese-Americans after Pearl Harbor than that given to persons of German and Italian ancestry.

Although the evidence we have so far assembled and will present here, does not unequivocally establish any of these hypotheses as stated, it does raise troubling questions about what happened in 1941 and 1942, and perhaps even more disturbing, how the Bureau has dealt with the issue over the intervening decades. As discussed in Section VI below, this evidence also suggests a series of improvements and safeguards for the future as well as needed future research.

Stepping back from these narrowly-worded hypotheses, we shall examine the Bureau's involvement in the internment process in terms of three broad topics: (1) the operational involvement of Bureau outputs, materials, expertise, and staff in planning and carrying out the internment; (2) the accounts provided by the Bureau, its staff, and its Directors concerning these events; and (3) the role of the racial classification and the available data by race in contributing to the internment of the Japanese Americans.

(1) The operational involvement of the Bureau

We divide our attention to this broad topic into four categories: (a) the use of macrodata, that is, the use of census results in terms of large aggregates and geographic units; (b) the use of mesodata, that is, the use of census results for very small geographic units; (c) the use of unprotected microdata, that is, census information that permits the identification of individuals; and (d) the use of other material, staff, services, and expertise provided by the Bureau.

(a) Macrodata

On Sunday, December 7, 1941 the Japanese armed forces attacked Pearl Harbor in Hawaii precipitating the United States entry into World War II. On Tuesday, December 9, 1941 the Bureau of the Census published its first report on Japanese Americans based on the 1940 Census, "Japanese population of the United States, its territories and possessions," followed immediately by reports on the "Japanese population by nativity and citizenship in selected cities of the United States" on December 10, and "Japanese population in the Pacific Coast States by sex, nativity and citizenship, by counties" on December 11 [US Census Bureau, 1974: items 1286, 1287, and 1288, respectively]. Additional preliminary 1940 census reports on the Japanese Americans were issued on December 19 and February 2, 1942 [US Census Bureau, 1974: items 1291, 1292, and 1293]. On December 12, 1941 the Bureau also produced a report on foreign born Germans and Italians in selected cities [item 1289] and only on September 30, 1942 a report on "Foreign white stock of German and Italian origin" [item 1299].

Copies of these reports can be found in some Census depository libraries and some were quoted in the press shortly after their release [NY Times, 12/13/41, p. 15; 12/30/41, p. 6; San Francisco Chronicle, 12/17/41, p. 15; San Francisco Examiner, 12/16/41, p. B; 12/17/41, p. 5; 12/24/41, p. A]. (Some of these numbers were referred to in the internal policy discussions in the period that preceded the internment decision, see section II.3 below.) In early January 1942 the Census Director, J.C. Capt, attributed the quick production of most of these tabulations to the fact that "we didn't wait for the declaration of war [which took place on Monday afternoon]. On Monday morning we put our people to work on the Japanese thing. And before the German and Italians hopped in we were working on that..." [Census Advisory Committee, January 1942: 20].

The production of census tabulations of this type would generally not be considered violations of the census confidentiality provisions. It should be noted that over and above the early start of work noted by Director Capt, the rapid tabulation of the data on Japanese Americans can be attributed to the fact that number of Japanese Americans was relatively small (under 300,000 nationwide in the 1940 Census) and, more importantly, because they could be identified by a single sort on race which has always been a 100 percent item in US census tradition. The possible role that the early availability of these tabulations may have played in either contributing to the growing fear of Japanese Americans on the West Coast or helping the Western Defense Command to form its conception of the internment effort will be discussed subsequently.

(b) Mesodata

As used here the term mesodata refers to statistical results presented at such a fine level of geographic disagregation that the results may be used in conducting field operations at the local level. The borderline between macrodata and mesodata is not always clear cut. In part it depends on the size of the geographic units, in part on the distribution of the target population among these units, and in part on the intended operational uses. For example, the special reports issued by the Census Bureau on December 10, 11, and 19, 1941 providing data on the number of Japanese Americans in specific cities and counties certainly provided a measure of operational assistance to those trying to locate the whereabouts of individual Japanese Americans. However, for those cities and counties that were comparatively large or that contained many Japanese Americans the operational significance of county-level or city-level data was limited.

The Census Bureau geography at the time of the 1940 Census was also capable of making finer geographic distinctions. These included entities, such as towns, townships, villages, or city wards, often collectively termed minor civil divisions. Another geographic unit often used in urban areas was the census tract, a well-defined and well-mapped way of dividing the land and population of a city. In 1940 each tract usually had a population in the 4,000-8,000 range [Truesdell, 1941: 365]. At these finer levels of disagregation the operational value of mesodata in locating a target population was clear to both the users and producers of 1940 Census mesodata on Japanese Americans. For example, during the January 1942 Census Advisory Committee meeting the following colloquy took place between Dr. Leon Truesdell, Bureau's chief population statistician, Dr. Virgil Reed, the Bureau's assistant director, and Director Capt:

Dr. Truesdell: ... We got a request yesterday, for example, from one of the Navy officers in Los Angeles, wanting figures in more or less geographic detail for the Japanese residents in Los Angeles, and we are getting that out....

Dr. Reed: [Commenting on all the hard work occasioned by numerous requests for data on the Japanese, Germans, and Italians] ... and some of them wanted them by much finer divisions than States and cities; some of them wanted, I believe several of them, them by census tract even.

Dr. Truesdell: That Los Angeles request I just referred to asked for census tracts.
...
The Director: We think it is pretty valuable. Those who got it thought they were pretty valuable. That is, if they knew there were 801 Japs in a community and only found 800 of them, then they have something to check up on....[Census Advisory Committee, January 1942: 20-21]

In fact, the Bureau had two geographic units even more compact than the census tract: enumeration districts and in some cities, census blocks and the 1940 census geocoding permitted tabulations at these levels of detail [Truesdell, 1941: 367]. Indeed, Dedrick [1981b: 172-173] testified to the Commission on Wartime Relocation and Internment of Civilians that at the end of February 1942 when he was on "detail to the War Department" the Census Bureau was asked to provide him,

a detailed cross-tabulation for even the most minute areas, the smallest areas for which data were collected. In other words, enumeration districts and in some instances cities by blocks.... Sheets of paper from the tabulation machines were sent out to WCCA at the hotel in Market Street in San Francisco, and became the basis for the WCCA statistical activities.

These recollections appear to be confirmed by contemporary Bureau records that included in one of its listings of war related activities for 1942, the provision of "photographic and photostatic copies of block maps for 10 cities" to the Western Defense Command [US Census Bureau, 1942b, Exhibit A:12]. That such block level tabulations were made and used operationally is also consistent with the 1972 recollections of Tom Clark, then serving as a liaison officer for the Justice Department with the Western Defense Command. Referring to the Bureau's help, Clark recalled "... They would lay out on tables [maps of] various city blocks where Japanese lived and they would tell me how many were living in each block" [US Commission on Wartime Relocation and Internment of Civilians, 1982: 105] (see Clark [1972] for a more complete text).

The usefulness of mesodata displayed cartographically can also be seen from figure 6 of the final report on the evacuation submitted by General DeWitt to the Army Chief of Staff [US Army. Western Defense Command and Fourth Army. 1943: 83] which reproduces a set of five population density maps of the "Japanese Population - Western Defense Command Area: 1940." The figure cites the "US Bureau of the Census" as the source and carries the legend, "each dot represents 10 people." As a precursor to a forced migration with a far more sinister ending, similar so-called "dot maps" prepared by the statistical authorities in Amsterdam played a major role in Nazi-inspired attacks in February 1941 aimed at several Jewish residential areas of that city [Seltzer, 1998: 525].

Recognizing the possible threats to the public arising from certain kinds of mesodata, the Census Bureau has progressively introduced stricter disclosure standards. Furthermore, beginning with the 1990 Census some deliberate noise has also been introduced into small area tabulations in an effort to provide further protection. Indeed, the Bureau has indicated, under the standards now in place the release of mesodata from the 1940 Census on Japanese Americans would have been severely restricted. Ironically, even as it was releasing mesodata from the 1940 Census pertaining to Japanese Americans, the 1940 Census marked the first time the Bureau introduced procedures to guard against inadvertent respondent disclosure from small area tabulations [Bohme and Pemberton, 1991: 18].

(c) Unprotected microdata

Whatever ambiguity may have existed in 1941 and 1942 about the propriety of using mesodata to target population subgroups (see, for example, Okamura [1981: 113-115] and Barabba [1981], quoted in part in Okamura), all agree that disclosure of data on specific individuals gathered under the protection of Title 13, the census act, was prohibited. In addition to statutory provisions, assurances made to the public in the years prior to the 1940 Census were clear and unwavering. For example, Corcoran [1963: 36] quotes President Hoover's 1929 proclamation for the 15th Census, including the explicit language that

the sole purpose of the census is to secure general statistical information regarding the population and resources of the country ... No person can be harmed in any way by providing the information required. The census has nothing to do with ... the enforcement of any national, state, or local law or ordinance. There need be no fear that any disclosure will be made regarding any individual person or his affairs ...

In response to a series of columns by Arthur Krock in the New York Times written in February 1940 expressing various concerns about the forthcoming census, including the possible misuse by different government agencies of information provided by the responding public, the then Census Director wrote a lengthy refutation which at one point quoted from the provision of Title 13 "that in no case shall information furnished under the authority of this Act be used to the detriment of the person or persons to whom such information relates" and at another point ridiculed the idea that census officials could be ordered to ignore the confidentiality provisions of Title 13 [Austin, 1940]. Finally, the 1940 Census enumeration form itself contained the reassuring words [US Census Bureau, 1979: 58-60] that

... Only sworn census employees will see your statements. Data collected will be used solely for preparing statistical information concerning the Nation's population, resources, and business activities. Your Census Reports Cannot Be Used for Purposes of Taxation, Regulation, or Investigation. [Capitalization in the original]

Despite this clearly articulated policy, rumors have persisted that names and addresses of Japanese Americans from the 1940 Census were provided to the military authorities to assist in the internment program. In a few instances, these rumors turned into concrete charges or new evidence emerged that seemed to confirm one or another aspect of these rumors. For example, in a 1982 book by John Toland, the author recounts a story based on a letter written in 1980 by one Henry Field to the effect that he (Field) was given the assignment by President Roosevelt on November 26,1941 to obtain the names and addresses of all American- and foreign-born Japanese listed by state and locality based on the 1930 and 1940 Censuses. Toland provides a wealth of detail, ostensibly based on Field's account, describing this massive enterprise which was successfully completed on December 3 [Toland, 1982: 268-269; 284-285; 351]. (It is not clear whether all the detail was Field's or some of it was added by Toland.) In any case, as it appeared in Toland's book, the tale contained so many factual errors and implausibilities (for example, the location of the Bureau's office, Director Capt's appearance and personality, the nature of his intercom, the length of time it would take to compile a roster of names and addresses from the census schedules, the number of pages of paper needed for such a roster), that it was effectively discredited by the Bureau shortly after it appeared [Chapman, 1982; US Census Bureau, 1982].

One source of misunderstanding was the remark by Tom Clark's in his oral history interview cited above, and referred to by Okamura [1981: 112-113], that "...the Census Bureau moved out its raw files" [Clark, 1972]. However, that remark seemed to refer directly to the block-level tabulations and maps provided by the Bureau from the 1940 Census and not to the availability individual names and addresses. Another example, also cited Okamura [1981: 115], was a July 1942 memorandum to Calvert Dedrick, seeking his help in locating the current addresses for two different groups of Japanese Americans [Bendetsen, 1942a]. This memorandum, which those writing on behalf of the Bureau seem to have ignored, has been dismissed by Daniels [1982] as irrelevant. Certainly, in and of itself, the memorandum may be explained in many ways that have nothing to do with the use of name and address information from the 1940 Census. As has been pointed out by Daniels [1982] and several of those speaking on behalf of the Bureau [Barabba, 1980; Chapman, 1982], initially the rumors may have been sparked over a failure to distinguish between the use of very detailed mesodata and the use of unprotected microdata.

During the course of its 1982 investigation of the Field/Toland charges the Census Bureau also indicated it had made a search of the archival materials pertaining to the Bureau stored in Record Group 29 maintained in the National Archives. On the basis of this review, the Bureau formulated its conclusions in two different ways: first, that it could find no records that would confirm that any disclosures of names and addresses from the 1940 Census had taken place, and second, that "the Census Bureau Director did not release any names and addresses from his census records to anyone at any time for any purpose" [Census Bureau, 1982: 1].

On the basis of our own review of what we considered to be the most relevant portions of Record Group 29, we too are able to conclude that the archival record does not establish that unauthorized disclosure of microdata took place. Unfortunately, we cannot agree that the archival record supports the Bureau's second formulation. At the same January 1942 Census Advisory Committee Meeting, in which Director Capt and his senior staff indicated that by January 10 that they were already providing tract level data on the Japanese Americans to the military, Director Capt expressed his willingness to take the next step:

We're by law required to keep confidential information by individuals. But in the end, [i]f the defense authorities found 200 Japs missing and they wanted the names of the Japs in that area, I would give them further means of checking individuals. [Census Advisory Committee, January 1942: 21]

The only comment offered was that by the chair, Dr. Murray Benedict, who noted briefly "I would guess you may have additional minor jobs in that," before turning to other matters.

Was Director Capt asked by the military for individual names and, if so, did he "give them further means of checking individuals?" The record at this point becomes murky. As described in the next subsection, within six weeks Dr. Dedrick, who participated in the January 1942 meeting of the Census Advisory Committee, was sent to San Francisco to assist Western Defense Command first in alien registration and subsequently in the detailed planning of the forced evacuation of the Japanese Americans. Furthermore, as already discussed, one memorandum has been found in which Dr. Dedrick's military supervisor does ask his help in providing names and addresses [Bendetsen, 1942a]. Equally disquieting are the inconsistent accounts of the Bureau's role (section II.2 below).

Given Tom Clark's detailed involvement in the planning operations for the internment and his description of the use of census maps and tabulations at the block level coupled with his omission of any mention of the use of names [Clark, 1972], it seems unlikely that names were released en masse by the Bureau. Moreover, there was no real need for this type of microdata since address information was effectively provided through the extensive release of detailed mesodata. On the other hand, one may speculate that Dedrick's presence on the West Coast provided a means for the Bureau to release confidential microdata on a selective basis in response to stated "military needs." The identification information on the punch cards included a reference to the schedule page and line number for each person. While as the Bureau has correctly pointed out searching through the original census schedules for all Japanese Americans would have been a massive clerical undertaking that would have been difficult to conceal [US Census Bureau, 1982], discrete searches for a few names might have taken place unremembered. It must be stressed that the evidence that any microdata from the 1940 Census were released in this manner is at this point conjectural. On the other hand, the fact that many searches of the 1940 Census schedules were carried out at this time for other purposes was acknowledged by the Census Bureau in its report for the fiscal year ending June 30, 1942 [1942a: 26]:

When information is desired about any individual person, the first step.... in referring to the schedule is to determine in what enumeration district that person lived at the time of the census. During the past year 500,000 addresses were allocated to enumeration districts preparatory to referring to the population schedules. Most of these searches were made for personal census record transcripts but a large number of these cases involved segregating information for special tabulations.

(d) The use of other material, staff, services, and expertise provided by the Bureau

In addition to the various types of outputs from the 1940 Census reviewed in the previous subsections, the principle assistance provided by the Census Bureau to support the internment program was the technical support and services provided to the Western Defense Command and the War Department. Central to this effort was Dr Calvert Dedrick. Dedrick served in San Francisco for about a year starting from February 27, 1942. In 1946 Dedrick described the circumstances of his involvement in these terms

The request for the assistance of the Census Bureau was made by General Allan Gullion, Provost Marshall General, in successive telephone conversation with Mr. South Trimble, Jr., Solicitor of the Department of Commerce, and Mr. J.C. Capt. At the outset it was anticipated that I would remain on the west coast for only 10 days or two weeks to assist Lt. Gen. John L. DeWitt in preparing forms and procedures for the registration of enemy aliens in that area. [Dedrick, 1946]

In March, after General DeWitt decided that evacuation of all Japanese Americans from the area was necessary, "arrangements were made by the Office of the Provost Marshall General with the Bureau ... for the loan of Mr. Dedrick to assist in this evacuation program" [1946].

On the West Coast, Dedrick worked under the direct supervision of Colonel Karl R. Bendetsen, Director of the WCCA and "recruited, trained, and operated a civilian statistical branch of that agency from March 1942 to March 1943" [1946]. On his arrival and until June 7,1942 Dedrick remained a Census Bureau staff member, thereafter at the War Department's insistence, he was transferred to the War Department payroll. In March 1943, he returned to Washington, DC where he worked in the Office of the Provost General on a variety topics, including "problems of alien evacuation and control" [1946].

The recruitment of qualified professionals to staff the statistical unit involved some effort. In 1981 Dedrick recalled that "my sociology friends [i.e., Dorothy Thomas, among others] over across the bay at Berkeley were very upset at me, but I must say I reached up to the University of Washington and got a Professor of Sociology [i.e., Calvin Schmid] and his principal graduate student [i.e., David Carpenter] as my right and left hand in this thing" [Dedrick, 1981b: 186-187].

The operational responsibilities of the WCCA statistical branch, headed by Dedrick, included

preparation of all necessary forms and procedures for the registration, temporary detention in Assembly Centers, and transfer to War Relocation Centers of all Japanese (sic) evacuees; the maintenance of adequate individual records through which the movement of each individual and each family would be traced; the maintenance of certain intelligence type records concerning individuals; drafting the details of the logistics of the evacuation movement including preparation of maps of individual evacuation area units by groups (of units) for temporary detention in Assembly Centers, and the regrouping of population for transfer to Relocation Centers; and for the preparation of tables, charts, and maps needed for over-all planning, operation and report purposes [Dedrick, 1946].

These are an unusually broad set of responsibilities for a statistical unit, at least in a democratic society. It is particularly surprising, both from the perspective of the Army and that of the Census Bureau, that Dedrick would be permitted, as the "Census Bureau's West Coast representative" [Capt, 1942b], to take charge of a number of the listed responsibilities. (Actually the precedent established in the WCCA for the statistical operation to serve as an administrative home for intelligence and control activities carried over into the WRA. )
In addition, the Population Division, the Division of Vital Statistics, and the Machine Tabulation Division were frequently called upon by WCCA and the WRA to provide special population estimates and special tabulations of Census Bureau data files as well as numerous tabulations based on administrative data files generated by the WCCA and WRA operations. A quick glance at the tabular material contained in many reports issued by the Western Defense Command, the WCCA, the War Relocation Authority, reveals the guiding hand of the Bureau's Manual of Tabular Presentation (see, for example, US Army [1943] and US War Relocation Authority [1946b]).

(2) The varying accounts provided by the Bureau, its staff, and its Directors concerning these events

Anyone attempting to reconstruct the extent and nature of the Census Bureau's involvement in the internment process faces three main obstacles: (1) with time, there are fewer and fewer eyewitnesses to the event and memories become less reliable; (2) as the Bureau [1982:1] has stated "... it could be surmised that if Census Bureau officials were secretly engaged in an action, including the violation of a federal statute, they would leave no trace of their actions;" and (3) the accounts of its involvement in the internment program have contained numerous substantial inconsistencies beginning with Calvert Dedrick's first memorandum on the subject [Dedrick, 1946].

Dedrick's immediate post war statement was written in July 1946, a year after he returned to the Census Bureau, for the Bureau's War History project. It contained two seemingly inaccurate assertions. First, in describing the use of 1940 Census punch cards and special tabulations, he observed, "However, the name and the individual identification data from the 1940 Census for the Japanese were not provided to the War Department nor were such data requested" [Dedrick, 1946]. As we shall see, many subsequent statements by the Bureau, including one directly attributed to Dedrick, assert that the War Department did make such a request, but that the Bureau refused. Second, in referring the short-term consultancy mission to the West Coast of a Bureau expert, Dr. Forrest Linder, he described Linder's assignment as working on "alien registration procedures." Four years earlier, as we discuss more fully below, he had referred to Linder's assignment under the heading "plans for the general registration of the civilian population" [Dedrick, 1942; see also Linder, 1942 and Dobbin, 1946b: 25]. Whether by chance or design, both of these errors served to minimize the Bureau's role in questionable activities .

Even in the final report of the evacuation prepared in 1943 by General DeWitt and his staff, one may observe a similar example of selective minimization with regard to the Bureau's role. While the 1940 Census was identified as the "most important single source of information prior to the evacuation" and the Bureau was given full credit for running special tabulations that "became the basis for the general evacuation and relocation plan," the actual tabulations referred to were "the aggregate Japanese population of states, the larger cities and groups of counties" [US Army. Western Defense Command and Fourth Army, 1943: 352]. There was no mention here, or elsewhere in the report, of the special small area tabulations of the 1940 Census used in operational planning or to generate the population density maps also shown in the report [1943: 83]. At the time this report was prepared, Dedrick was still on the West Coast working for Bendetsen and his help with this report has been noted [Dobbin, 1946a; Dedrick, 1981b: 178].

About the same time, in the Bureau's own pamphlet describing itself to the war-time American public [US Census Bureau, 1943a: 7-8], referred to its use of the 1940 Census data in sub-section, "War Facts Were Ready" to grind "out promptly and in detail" answers to such questions as

How many alien Japanese and Germans and Italians have we, and where? How many naturalized, and second generation? How old are they, and what are their occupations in specific areas? How thick are they in strategic areas and possessions, such as Alaska and Hawaii?

There was no further mention as to how these data were used or produced, except for the captions to several illustrations extolling the power of punch cards.

The discussion of the use of 1940 Census data in the chapter of the unpublished war history volume dealing with the Bureau's role in the evacuation repeats verbatim the statement in Dedrick's memorandum that name and individual identification data from the 1940 Census were neither provided to the War Department nor sought by them [Dobbin, 1946a: 4]. It then goes onto add

This was in conformance with Census law, under which the Bureau must keep such data confidential, excepting that it can give these data to an individual when he requests it. [1946a: 4]

A year after the war history volume was completed, when the Census Bureau issued its public information pamphlet, Fact Finder for the Nation, all mention of the Japanese Americans had vanished and the brief summary of the Bureau's war-time contributions focused elsewhere, "the marshaling of facts on materials and manpower during Word War II was one of the important aspects of accomplishing the miracle of production that made possible the winning of the war." [US Census Bureau, 1947: 4]. The omission cannot be attributable to staff turn-over since when the pamphlet was issued, Capt was still Director and Dedrick was still a member of the senior management team.

An extended period of institutional silence then ensued. When the subject of the Japanese Americans and the 1940 Census subsequently reemerged, the Bureau's role was cast in far different terms. In marked contrast with both its earlier accounts and its long silence, at some point in the 1960's or early 1970's the Bureau's war-time involvement in the internment program began to be cited as an example of how the Bureau was able to protect census respondents from the efforts of others in Government to harm them.

In time, this version of the story became quite specific. For example, in 1975 the then chief of the Census Bureau's history staff, responding to a question by the historian Roger Daniels, wrote [Bohme, 1975]:
According to Dr. Calvert L. Dedrick, who was Chief of the Census Bureau's Statistical Research Division in 1942, "someone" in the office of the U.S. Army Provost Marshal General telephoned Mr. South Trimble, Solicitor of the Department of Commerce in Washington, on or about February 24, 1942, and asked for detailed information from the 1940 Census records, culminating in a list of individual Japanese Americans by name and address. The Solicitor's response was that no names or addresses would be disclosed, but that the Census Bureau could prepare a special tabulation of persons of Japanese ancestry by minor civil division (which the Bureau subsequently did). There is no correspondence involved in this particular interchange.

As we established in section II.1.b above, census tract data from the 1940 Census had already been requested and provided in early January 1942. According to Daniels [1982: 105], in 1975 Dedrick declined an opportunity to comment further on this account by Bohme. Subsequently Dedrick apparently returned to his 1946 position, stating that "at no time were the names and addresses of individuals or families received or requested from the Bureau of the Census" [Dedrick, 1981a: 4].

By early 1980, the Bureau's description of the episode went even further in articulating the heroic lengths it went to protect Japanese Americans. As part of its preparatory program for the 1980 Census, the Bureau launched an effort to involve college and university students in the enumeration process and issued a textbook to assist in their training [US Census Bureau, 1980]. That work, in a chapter entitled "Facing Some Issues," included in its review of the Bureau's track record in dealing with privacy and confidentiality issues the following account [1980: 76-77]

The Census Bureau was even able to resist intense political pressure for the release of data. For example in 1941, with World War II underway, there was near hysteria about the fact that there were many Japanese-Americans living on the West Coast. This led to one of the most embarrassing moments in U.S. history - the confinement of large numbers of loyal Japanese Americans. At the height of this feeling, the Secretary of War is believed to have requested the Census Bureau to supply, the names, addresses, and ages of all persons of Japanese extraction living on the West Coast. This time, in spite of the national emergency, the Bureau held to its position on confidentiality of individual records and refused to release the information. The Bureau did supply summary data for political jurisdictions, but no individual data were released.

Okamura [1981: 111-112 and footnote 2] cites several other sources where virtually the same retelling of the story is used in the mass media, including a television script provided Michi Weglyn perhaps dating back to 1960. In the same article, Okamura went on to question the Census Bureau's depiction of itself as defenders of the Japanese Americans in the post Pearl Harbor period, citing several documents [for example, US Army, 1943; Bendetsen, 1942a; Clark, 1972], already discussed here. He concluded "there is no doubt that the Census Bureau was intimately involved in the planning for the mass incarceration of Japanese Americans. The only question is whether their actions violated the law" [Okamura, 1981: 113]. He then quoted [1981: 113-114] from a letter from then Census Bureau Director, Vince Barabba that attempted to respond to the information that he and Weglyn had unearthed.

In the quoted portions of his letter, after first acknowledging that "a different account would be more appropriate in the future," Barabba [1980] basically made two points in defense of the Bureau: first, that the confidentiality law that governed the 1940 Census did not extend to any harm that arose from tabulated data; and second, that the War Powers Act of 1942 suspended the confidentiality provisions of the census act. Although he indicated that "we have no evidence ... that identifiable confidential information was ever released during this period because of the [War Powers] act," Director Barabba's references to the War Powers Act actually raised more questions than it answered. As we have seen, Director Capt as far back as January 10 had expressed a willingness to provide the defense authorities with confidential information from the 1940 Census in violation of his understanding of the law [Census Advisory Committee, January 1942]. This occurred some two weeks prior to January 22, 1942, the date that the bill that eventually became the War Powers Act was introduced. Moreover, according to Dedrick's 1975 account, a request for names and addresses from the 1940 Census was made by the War Department to the Department of Commerce on or about February 24, 1942 [Bohme, 1975]. Under the then pending legislation for the War Powers Act, requests for otherwise confidential information held by any bureau in the Department of Commerce, were to be made through the Secretary of Commerce. Thus, the request if made in February 1942 could have lawfully been provided in a month's time under the provisions' of the War Powers Act. The question then arises, what led Director Barabba to refer to this Act?

In 1982 Census Director Chapman, again wrote to Okamura, forwarding and endorsing a five- page report responding to the charges in the Toland book [1982]. This report [US Census Bureau, 1982], in addition to refuting Toland's charges, reiterates and elaborates several earlier statements made by the Bureau about its involvement in the internment. These include: (1) a description of the macrodata provided that implies that the tabulation runs for "Japanese, German, and Italian Americans" were equivalent and obfuscates the timing of the post-Pearl Harbor publications [1982: 2]; (2) a recollection by Dedrick that "he received from the Census Bureau, early in 1942, aggregate data on the Japanese American population by citizenship status and by place and county subdivision within each state on the West Coast" [1982: 3]; and (3) another recollection by Dedrick "that about the time of his departure for San Francisco, on February 25, 1942, Director Capt told him that inquiries from War Department staff as to the use of the names and addresses from the population schedules had been refused but that tabulations of data for small areas could be supplied" [1982: 3].

In 1991, two members of the Census Bureau historical staff, indicating that they were reporting "the general results of research," summarized the Bureau's role in the internment using the following language [Bohme and Pemberton, 1991: 12]:

the Bureau turned down the War Department's telephoned request in 1942 for the names and addresses of Japanese (sic) living in Western States. What the Census Bureau did do, however, was to give the War Department stacks of punch cards (which had no names or addresses) identifying such persons by census tract or other small area - information readily to available to anyone in the published reports, but in more convenient form. The military authorities thus knew where to concentrate their efforts to intern these people, but in no case did the Census Bureau, contrary to law, furnish information about individuals that could be used to their detriment.

While some tract and block data from the 1940 Census were eventually published by the Census Bureau, although not in early 1942, data for individual enumeration districts as such were never made available to the public.

In 1995 former Census Director Bryant, largely relying Bohme and Pemberton's account, even as she began to distance herself from it, wrote that after Pearl Harbor

The War Department requested the names and addresses of all Japanese and Japanese-Americans enumerated in the 1940 Census in the Western states. The Census Bureau rejected the request. But the Bureau did give the War Department punch cards, without names or addresses, identifying census tracts and other small areas with high concentrations of Japanese.... Bureau defenders contend that the population counts provided to the military were already available in published reports. Even today, the Census Bureau promises confidentiality only to individuals. Neighborhood characteristics... are available to all.

The Japanese-Americans who lost property and were interned simply because of their ancestry see things differently, indeed. Their devastating fate has been cited repeatedly by critics as a breach in spirit, if not in fact, of census confidentiality. [Bryant and Dunn, 1995: 32-33]

Unfortunately, during the period leading up to the 2000 Census, the image of the Bureau's protection of the Japanese Americans early in Word War II was once again invoked to reassure respondents more generally. For example, earlier this year the Chicago Tribune [Ibata, 2000] reported that

Michael Voss, manager of the Census Bureau's Palatine office, urged executives to assure their employees that census data is confidential .... "During World War II, the president tried to obtain information from the Census Bureau about Japanese Americans, and the President himself was turned down," Voss said.

Since we have heard of this same example being cited by local census officials elsewhere in the country (see, for example, the Bergen (NJ) Record [2/9/00, "Ducking the Census," opinion section, p. 8]), we can only assume that the example was, as in the 1980 Census, included in the training materials for the present census. This despite Director Barabba's 1980 disavowal of this sort of characterization of the Bureau's behavior and Director Bryant's more recent comments. The only change in the story from that used in 1980 is that now the Bureau is said to have turned down the President rather than merely the War Department. (It should be noted that as soon as we became aware of this problem in the 2000 Census, we notified the Bureau, and they took steps to correct the record in so far as possible.)

Not only was the Census Bureau disseminating a falsely romantic myth about its past behavior to the public, it appears that this myth has been internalized widely among the Bureau's permanent professional staff. During informal discussions over the past year with a range of staff members, we have repeatedly heard that the story of the heroic steps taken by the Census Bureau to protect the interests of Japanese Americans respondents at the outbreak of World War II is accepted as fact within the Census Bureau.

The numerous accounts of the use of 1940 Census data to assist in the internment of Japanese Americans provided by the Census Bureau, Census staff, Census directors, and others, over the years, vary so greatly that the scope of the problem may become lost in the detail. By way of clarification, Table 1 presents summary information on critical aspects the Bureau's involvement drawn from the different accounts discussed in this section. A glance down each column of the table reveals such inconsistencies and departures from the truth, whatever their cause, that taken together seriously undermines the Bureau's credibility. This is particularly unfortunate, given that at the outset of Bureau's work on the War History Project that the stated goal was to include

not only a record of what we did, but why we did it, and how, and how it could be better done if we had to it over again. There is no point in simply making it a boastful record of achievement. We should record failures as well as success [Capt, no date].

The Bureau has a long tradition of respecting the truth of this view with respect to various types of statistical errors. However, the proposition is equally apt with respect to policy and administrative mistakes.

Table 1. Summary of Available Information on Aspects of the Use of the 1940 Census
in the Internment of Japanese Americans

Reference(Source and date)(a) Date information requested or provided(b) Lowest level for whichinformation providedand nature of information(c) Was the Bureau asked to provide names and addresses?(d) Census Bureau response(e)
US Census Bureau, 1941c 12/11/41 County - statistics Not applicable Not applicable
Census Advisory Committee, January 1942, pp. 20-21 1/10/42 and earlier Tract - statistics Not mentioned Director Capt: "If the defense authorities ... wanted the names ... I would give them further means of checking individuals."
US Census Bureau, 1942, p. 12 1942 Block - maps Not applicable Not applicable
Bendetsen, 1942a 7/3/42 Dedrick requested to provide addresses of individuals Possiblya Not known
US Army, 1943, p. 352 (Drafted by Dedrick) Early 1942 "States, larger cities and groups of counties" Not mentioned Not applicable
Dedrick, 1946 Not applicable Not applicable "Not requested" "Not provided"
Dobbin, 1946a Not applicable Not applicable "Not requested" "Not provided"
Clark, 1972 Prior to 3/42 Block - statistics Not mentioned Not applicable
(a) (b) (c) (d) (e)
Bohme, 1975, citing Dedrick On or about 2/24/42 "Minor civil divisions" Yes, by the War Department "Names not provided"
US Census Bureau, 1980, pp. 76-77 (Drafted by Bohme) Unspecified "Summary data for political jurisdictions" Yes, by the WarDepartment "Names not provided"
Dedrick, 1981 Just after 2/26/42 Detailed cross tabulations for blocks and enumeration districts "Not requested" "Not provided"
Bohme and Pemberton, 1991, p. 12 In 1942 Punch cards identifying people, but without names and addresses, by census tract and other small area Yes, by the War Department Names and addresses not provided
Ibata, 2000, citing a local Census Bureau official in Illinois During World War II Not mentioned Yes, by the President Information not provided

(a) It is unclear whether this request to Dedrick for addresses of specific individuals involved
the use of 1940 Census information or what Dedrick's response was.

(3) The role of the racial classification and racial data in contributing to the internment of the Japanese Americans

Regardless of the extent to which data from the 1940 Census was used by the Western Defense Command in the initial operational planning of the internment program, did the availability of data on the Japanese Americans in the 1940 Census influence the decision to launch the internment program or at least the form it took? If such an influence can be identified, how did it seem to operate? These questions are particularly relevant because the war-time experience of the Japanese Americans was so radically different that of German Americans and Italian Americans. In addition, the treatment of the concept "Japanese" by the census and the US Census Bureau also differed radically from that used for the other groups. That is, while the concepts "German" and "Italian" are used solely in the census to refer to countries of birth, the concept "Japanese" is used both to denote a country of birth and as a racial category. In short, were the differences in census treatment in any way related to the differences in how these different population groups were treated during World War II?

In order to examine the possible influence of the existence of Japanese as a racial category in the 1940 and earlier censuses on the treatment of Japanese Americans during World War II, we posit three modes of influence: (1) influence operating directly on the Western Defense Command or on the War Department in Washington, DC, between December 7, 1941 and March, 1942, by which time the major decisions had been taken; (2) influence operating through the perceptions of the general public and opinion leaders on the West Coast of the nature and scope of the 'Japanese problem' during the same period; and (3) influence operating through the continuous fueling of endemic racism. It is possible for more than one of these modes of influence to have played a role.

Without more thorough research than we have been able to so far carry out, we have not been able to fully document or quantify the importance of any of these influences. We have examined two of the major sets of published archival material relating to the internment decision process in the War Department and other concerned federal agencies in Washington, DC and in the Western Defense Command [Daniels, 1989 and the CWRIC, 1984]. Although we found numerous references to population counts and estimates of various sorts, including the invocation of figures from the 1940 Census in critical discussions within the Western Defense Command, these references often raise more questions than they answer. We realize, however, that we were constrained by the selections made by others from the full set of relevant materials in the archives. Accordingly, it is important to reexamine some of the basic archival sources to search out additional relevant material. Having done so, we could then reassess the influence of the census concept and the figures based on it using the larger set of materials related to the decision to intern all Japanese Americans and not just those who were foreign born or aliens.

Similarly, pilot research we carried out aimed at examining the influence of Census Bureau outputs and activities on public perceptions of the "Japanese" problem on the West coast in the months following Pearl Harbor produced suggestive, but very incomplete results. For example, searches for data-driven stories concerning Japanese Americans in two California newspapers in the two months following Pearl Harbor found roughly similar peaks and valleys in such coverage to those that had been found by Grodzins [1949, Appendix I]. (Grodzins based his study on all Japanese American related stories in a far-larger number of California newspapers.) Our preliminary research has also revealed that the Census Bureau did not simply tabulate and publish the special 1940 Census reports on the Japanese Americans. It also sent copies of at least one of these reports [US Census Bureau, 1941c] directly to concerned local officials, including the Governor of California [San Francisco Examiner, 12/17/41, p.5].

Further research in this area would attempt to replicate Grodzins' earlier study of West coast press coverage in terms of the sub-set of stories based on reports of data attributed to the Census Bureau. Such a study would document more fully how widely the Bureau results were picked up and the extent to which Grodzins' over-all conclusions on total press coverage were a function of data-based stories. Given the direct provision of data by the Census Bureau to local officials on the West Coast it would also be useful to re-examine the archival records of the relevant government offices to assess whether these data had any apparent impact on the formulation of local views and policies.

Whatever the outcome of such further research, it seems self-evident to us that influence of the continuing and unthinking use by the Census Bureau of a "racial/ethnic" classification that was first used to reflect the constitutional requirements of racially based slavery. Subsequently, the classification not only evolved in response to the shifting racist fashions of the 19th and early 20th centuries, but it also helped to foster a range of race-based thinking by the public. In other words, the racial classification used in the 1940 Census, including its choice of categories, was likely both a reflection of an endemic racism that led to the internment of the Japanese Americans and an important contributing factor that helped to continually fuel racism generally and shape the specific basis and geographic scope of the internment program.

Racist thinking with regard to persons of Asian ancestry tended to have a more direct impact on the Bureau's racial/ethnic classification over the years than for those of European ancestry. For example, while persons of Swedish and Italian ancestry stereotypically may differ in appearance, the census captured any differences in these populations through such items as country of birth or country of birth of parents. In the case of those of Asian ancestry, analogous national differences were (and still are) captured by the Census Bureau through the introduction of nationality-based racial categories into the race item, such as Chinese, Japanese, or Korean. Given this background, it is not surprising that on the morning of December 8, 1941 the decision was made at the Bureau to produce tabulations of Japanese Americans based on the item on race rather than on country of birth or citizenship. It also helps to explain why this decision was not altered in the following weeks after the Bureau produced tabulations of German Americans and Italian Americans.

As is also clear from material quoted earlier, senior Census Bureau staff and the members of the Census Advisory Committee, like most Americans at that time, constantly referred to all person of Japanese ancestry, whether living in Japan or the United States, as "Japs." Indeed, at one point during the January 1942 Census Advisory Committee meeting, Director Capt thought it necessary to remind those present that most of the "Japs" enumerated in the 1940 Census were American citizens.

III. Registration Systems for Military Service and Aliens

During World War II, military authorities claimed that the Japanese population on the West Coast posed a national security threat, and they justified the internment of the Japanese alien and citizen population on these grounds at the time. The CWRIC investigations and report and the subsequent actions of Congress in the 1980's repudiated that claim. Nevertheless it is widely recognized that nation states have an obligation to defend their own existence and the safety of their residents, and that in time of war, a government will make extraordinary demands on its population, including drafting people into war service, deploying a surveillance apparatus for national defense, and otherwise instituting control mechanisms, which would be abhorrent during peacetime. Hence it is still possible to propose that despite the repudiation of the decision to intern the Japanese population, the federal statistical system and the Census Bureau did not violate their own rules and procedures or do anything improper while participating in the administration of the internment.

Dealing with this claim involves suspending judgment on the legality and propriety of the internment for the time being and considering alternative scenarios for protecting the West Coast from the threat of invasion and domestic terrorism in December 1941 and early 1942. We may ask what other and related roles in national defense did federal statistical agencies play in this period? The answer takes us into the history of the population data systems that were used to draft men into military service and to control any potential threat posed by the large alien population in the United States. Unlike the census, these registration systems were primarily designed to serve administrative functions. They could serve research functions as we shall see, and senior officials from statistical agencies were involved in their development, but the administrative requirements of the registrations meant that they were designed to monitor individuals and their behavior. Men were to be drafted into the army and aliens were to be watched for signs of disloyalty and encouraged to naturalize and thereby prove their loyalty to the United States. Our research and conclusions on these matters are very preliminary, though they do provide, we hope, an important new framework for evaluating the role of the 1940 Census in the internment of the Japanese and for considering the relationship between administrative record systems and survey operations more generally.

The background for the development of these registration systems from the perspective of the statistical system lies in the larger initiatives of the New Deal. In the late 1930's the younger generation of statisticians and social scientists at the Census Bureau and related federal statistical agencies began to expand their horizons for research on the American population [Anderson, 1988]. Since the onset of the Depression, they had confronted what they saw as the weakness of federal population statistics and began to envision how they might gather current and continuous data to monitor the condition of the American population. The development of population sampling methods for use in consumer expenditure and unemployment surveys was one such avenue. The coming completion of the vital registration system for the entire country was another. Yet, at the time, any notion of developing a system to generate continuous national, population data seemed a very faint possibility indeed since there seemed to be little likelihood that Congress and the administration would ever fund an ongoing program of population statistics. Articles about the successful implementation of such comprehensive systems in Europe, particularly in the Netherlands, appeared in American demographic and statistical journals [Methorst, 1936, 1938; Thomas, 1937]. At best, American researchers had to hope they could attach proposals for data collection and research to new programs as they came along.

A major opportunity along these lines arose in 1937. After debating the need for a national unemployment census for years, Congress and the administration finally funded a national registration of the unemployed through the post office. John Biggers was put in charge of the effort and census officials Calvert Dedrick and Morris Hansen lent technical support. They convinced Biggers to include a sample check census administered by post office personnel along with the voluntary registration. The sample survey proved that the voluntary registration undercounted the unemployed. It was the first national area sample survey and its success prompted further efforts to introduce sampling into the 1940 Census design. The 1937 check census is generally considered the germ of the idea for the monthly unemployment survey, the Monthly Report on the Labor Force, now the Current Population Survey [Duncan and Shelton, 1978; Dedrick and Hansen, 1937].

This history of the origins of the innovations in statistical sampling methodology is generally known to students of the history of federal statistics. What is less well known is that the expertise developed during these data collections was not limited to the development of survey sampling. These efforts also gave key federal officials expertise in developing and assessing the quality of data from population-based registration systems of one kind or another, and the war years saw several major efforts in this regard.

In the late 1930's, the federal government turned its attention to its defense needs. War broke out in Europe and Asia, and Congress and the administration called for further measures, which would lead naturally to such population-based registration systems. In June 1940, the Alien Registration Act of 1940, also known as the Smith Act, required that all aliens register and be fingerprinted at their local post offices. In September 1940, the United States instituted its first peacetime draft, requiring all men between the ages of 21 and 35 to register for possible military service [Hutchinson, 1981: 540-542].

Both these registration efforts were very large projects. The alien registration encompassed 5 million aliens. The draft registration signed up 16.4 million men. For social researchers, these new laws signaled the possibility of building more complete and more sophisticated data systems than one could derive from the most ambitious existing population data collection, the decennial census. And they presented interesting technical challenges for questionnaire design, possibilities for record linkage and continuous updating. As the October 1940 issue of Population Index reported ["Current Items," vol. 6, no. 4 (Oct. 1940), pp. 250-1]:

A system of registering a large segment of the population of the United States was instituted by the Alien Registration Act. . . . Whatever the primary purpose of the Act, it may prove of importance to students of population by establishing a precedent for non-military registration as well as by providing statistics on non-citizens.

And the American Journal of Sociology ["News and Notes," Vol. 46, No. 3, Nov., 1940, p. 381] reported that

Calvert L. Dedrick, chief statistician, Division of Statistical Research in the Bureau of the Census, has been loaned by that agency to serve temporarily as a technical consultant to the solicitor-general of the Department of Justice, to advise on the solution of technical problems involved in the registration of aliens.... The act providing for the alien registration specified not only that all aliens must register but that registration be kept currently up to date, taking account of the movements of aliens within the United States as well as movements to the United States.

Both these systems involved the federal government in major new population-based registration systems that went far beyond any previous effort. It is useful to recall that although the United States had been a nation of immigrants since its inception, and had required ship captains to provide lists of immigrants arriving since the early nineteenth century, the federal government had given up efforts to maintain an alien registration system in 1828 [Neuman, 1996: 40-41]. The selective service system, instituted during World War I, was quickly dismantled during peacetime.

The administrative issues involved in developing questionnaires, retrieving the data, and determining where the records should be located and how the forms should be organized for both statistical and administrative purposes were formidable. From the perspective of the administrative agencies, the data were designed to facilitate a military draft and the monitoring of subversive activities in case of war. The statistical experts saw broader possibilities for the grounding of "American population policy" and the analysis of demographic trends (see e.g., Lorimer, Winston and Kiser [1940]). For both administrators and researchers, there was interest in getting the basic data collected properly and accurately. The experience of the unemployment registration indicated that the researchers could anticipate an undercount in the registration because of inexperience with the new, far-flung efforts, and because the purposes of the registrations were inherently controversial. Officials were aware that individuals might not want to admit to being unemployed. Individuals might try to avoid registration as an alien or for military service because of the social stigma of admitting one's status, or out of fear or suspicion of government. The alien registration was particularly problematic, and generated a storm of debate in the media about the loyalty of the foreign born, created worries over discrimination against immigrants, and widespread press coverage of the entire effort.

Census officials responded to these concerns by offering technical help in developing the registration methods. They also suggested using the population census to provide comparative data for evaluating the completeness of both the draft and alien registration efforts. The first offer facilitated the development of ties between key statisticians in the Census Bureau and those in the administrative agencies of the Justice and War Departments managing the registration programs. The second led to rather awkward questions about the accuracy of the census.

In June 1940, Calvert Dedrick reported to Leon Truesdell, head of the Census Population Division, that the War Department had requested estimates of the numbers of draft age men. Dedrick reported to Truesdell that he had the estimates prepared, and that the task was 'somewhat hazardous' -- presumably because the 1940 census was still in the field. [Dedrick, 1940a]. We know now that whatever undercount existed in the draft registration, the estimates the Census Bureau reported were also low. More men registered than the Bureau estimated, and this finding spurred further analysis of what became the issue of the differential undercount in the census [Anderson and Fienberg, 1999].

In the same memo, Dedrick also reported to Truesdell that he was "discussing the subject of alien registration" with staff in the Department of War, and in late July, he sent a copy of an alien registration form he had developed to Truesdell for comments [Dedrick, 1940b]. Alien registration was the responsibility of the Immigration and Naturalization Service of the Department of Justice (INS). The INS required basic estimates of the number of aliens likely to register in order to print the proper number of forms, deliver them to post offices, etc. The registration period ran from August to December 1940. Initially the INS estimated that 3.6 million aliens would register. By late fall, more aliens were registering than anticipated. The number eventually approached 5.0 million [NY Times, 7/6/40; 7/31/40; 9/2/40; 12/27/40; 2/20/41]. By December 1940, when the alien registration period ended, the INS was running behind in returning the receipt cards to the registrants; they estimated they would not finish mailing them until February 15, 1941.

With the completion of the registrations, the information was available for administrative purposes. For the alien registration, the FBI matched fingerprint files from the alien registration with their records of spying, crime and subversive activity in the spring of 1941. Records were stored centrally in Philadelphia, and at local FBI offices around the country. In 1940 and 1941 the FBI, Special Defense Unit of the Justice Department, the Office of Naval Intelligence and Army Military Intelligence (G-2) began compiling lists of potential subversives to be apprehended in case of war onto what came to be called the ABC lists [Irons, 1983]. Military and Justice department officials planned for internment camps and alien detention centers in case of war, and local police were encouraged to track potential subversives.

Once the U.S. entered the war on December 7, 1941, these surveillance systems were called upon to protect the nation from internal subversion and further external attack. The war declaration created a new category of 'aliens,' namely 'enemy aliens.' President Roosevelt's Proclamations of December 7 and 8, 1941 defined the status of enemy aliens from Germany, Italy and Japan, and specified that "all natives, citizens, denizens, or subjects of (those countries), being of the age of fourteen years and upward, who shall be in the United States and not actually naturalized, shall be liable to be apprehended, restrained, secured, and removed as alien enemies" [quoted in Daniels, 1997]. If apprehended, the alien was entitled to a hearing before an Enemy Alien Hearing Board, though the alien was not entitled to legal counsel. Daniels estimates that 11,000 enemy aliens were interned overall during World War II, though it is thought that many more aliens were arrested than interned.

It fell to the Justice Department, including the INS and the FBI, and the military, including its intelligence arms, to manage the administrative issues of interning enemy aliens. The Justice Department set up an Alien Enemy Control Unit on December 22, 1941 [Irons, 1983: 23], started arresting those on the ABC Lists, and began planning for a second registration activity, this time of 'enemy aliens' only. As we have already observed, military officials on the West Coast requested and obtained from the Census Bureau small area tabulations of 1940 Census data as well.

The story of the control of enemy aliens is a complex one and has been told elsewhere (for example, Irons [1983]; tenBroek, Barnhart and Matson [1954]; and CWRIC [1982]). Suffice it to say here that the Justice Department and military officials in Western Defense Command clashed over the rights of enemy aliens, including the need for individual warrants for searches, and appropriate travel and other restrictions on their activities. The war news in the Pacific was extremely bad in early 1942, and it was believed that Japanese submarines were operating off the West Coast. There was great fear of invasion or subversion, even though the FBI, the Office of Naval Intelligence and G-2 had no definite evidence of subversion. Nevertheless, civilian political leaders and military leaders, particularly General DeWitt of the Western Defense Command, feared that the 200,000 enemy aliens' households on the West Coast posed a greater national security threat than the Justice Department realized.

In late December and early January 1942, the War Department began to demand that the Justice Department accelerate the enemy alien registration, issue regulations for seizing contraband held by enemy aliens, and consider the creation of prohibited or restricted zones from which enemy aliens could be removed. In early January the Justice Department acceded to mass raids on the homes of enemy aliens who could be suspected of having illegal 'contraband,' including radios and other devices which could be used to signal the enemy, and scheduled the alien enemy registration for the week of February 2, 1942.


Good data on the location of enemy aliens was necessary for these programs, and the natural source was the 1940 alien registration records. We are not aware of any definitive research evaluating the role of the 1940 alien registration records in the control of enemy aliens generally, and on the West Coast in particular during the first months of 1942, and we have yet to examine the issue in depth. We do know that the FBI Agent in Charge in San Francisco, Nat Pieper, reported to Col. Forney of the Western Defense Command on January 1, 1942, that he found the

Immigration Records regarding enemy aliens. . . in a condition unsatisfactory for prompt use and estimates that extensive clerical work over a period of time will be necessary to put them in shape for ready use [Daniels, 1989: vol. 2, no page].

The raids in search of contraband started the first week of February 1942. The Justice Department also asked that as the Army and Navy identified "prohibited" and "restricted" zones for alien enemies [tenBroek, et al., 1954: 104], they also submit to the Justice Department "detailed plans for evacuation and resettlement." The Army began to submit lists of prohibited and restricted areas in late January 1942. As the number and size of the prohibited areas grew, the Justice Department began to object that the removal of enemy aliens from such large areas, for example, from the entire city of Los Angeles, would violate their rights and would be logistically impossible.

The simmering conflict came to a head in early February, and the Justice Department ceded authority for the control of enemy aliens to the War Department. Throughout late December and January 1942, calls appeared in the press and in the internal government memoranda for the arrest and removal of 'the Japanese' from the West Coast, not just alien Japanese. On February 19, 1942, the President issued Executive Order 9066, which authorized the Western Defense Command to remove Japanese American aliens and citizens from the West Coast.

The Wartime Civil Control Administration (WCCA) had the responsibility of managing the removal, and as we have already noted, the Final Report of the Japanese Evacuation from the West Coast, 1942 claimed that the "most important single source of information prior to the evacuation was the 1940 Census of Population" [US Army, 1943: 352]. Once the government expanded the enemy alien control program to include citizens of Japanese ancestry, the registration systems built for enemy alien control were no longer sufficient.

The "Statistical Summary" in the Final Report, written by Dedrick, makes the issue plain. "Consideration was given by the Wartime Civil Control Administration to the possibility of conducting a general compulsory registration of all enemy aliens and of native-born persons of Japanese ancestry prior to, and in preparation for, the evacuation program." The WCCA decided against such a registration for several reasons [US Army, 1943: 355]:

1. A registration would require some time for organization and tabulation before it would yield results useful in the evacuation program.
2. A registration of all enemy aliens had been conducted in February by the Department of Justice, and copies of the forms were on file with the local offices of the Federal Bureau of Investigation and at the Alien Registration Division, Immigration and Naturalization Service, Philadelphia, Penna. Certain preliminary data were made available to the Wartime Civil Control Administration from local counts of these registrations, but no detailed classified information, such as the distribution by small geographical units, age, sex, occupations, family size, etc., were available.
3. The Japanese community already anticipated evacuation, and it was felt doubtful that accurate information would be given in all cases.
4. A complete evacuation of the coastal area was contemplated, and a preliminary verification with Federal Bureau of Investigation and Military Intelligence data showed that the Census of 1940 would be sufficiently accurate for the controlled movement.

In other words, the 1940 Census data for Japanese Americans came to serve as the population registration system necessary for "the controlled movement."

IV. General Population Registration for Military and Statistical Purposes

As already recounted, Dedrick's primary duties in San Francisco were to provide on behalf of the Census Bureau his considerable technical and managerial skills to assist the military authorities in planning and implementing the internment of Japanese Americans on the West Coast. Nevertheless, he also fostered initial work on a broader effort that potentially effected all Americans - a scheme for registering every American. The specific background of this effort is not clear.

On the one hand, Col. Karl Bendetsen, then Assistant Chief of Staff, Civil Affairs Division, WCCA, has indicated that the initial work on the idea was carried out at his behest [Bendetsen, 1942b]. (The WCCA was the temporary military organization established under the Western Defense Command and Fourth Army to plan and carry out the evacuation of Japanese Americans. Bendetsen, who eventually became an Under Secretary of the Army, has been considered by many to be the person most responsible for turning the idea of internment into a reality.) On the other hand, just prior to his departure for San Francisco in late February 1942, Dedrick drafted a letter that Census Director Capt sent to Professor William K. Ogburn of The University of Chicago as the new Chair of the Census Advisory Committee that contained the following [Capt, 1942a]

There is a growing feeling that a general registration of the inhabitants of the United States may become necessary. The Bureau has given some thought to this problem and has assisted other agencies, most notably the Immigration and Naturalization Service, Alien Registration Division, in the preparation of forms and the working out of satisfactory office procedures....Should the legislative authority be given and sufficient funds made available, I believe the Census Bureau is in a position to assemble the data and keep a registration file for every inhabitant in the United States. There is no question but that we are better prepared to do this work than any other agency.

In any event, on May 9, 1942, Forrest E. Linder, a Census Bureau expert in the area of demographic statistics, completed a short mission to the West Coast for the WCCA and submitted a 35 page report, "Memorandum Relating to a General Population Identity Registration for Military Purposes," along with a brief covering memorandum to Dedrick [Linder, 1942]. At this point Dedrick was still a Census Bureau staff member and Chief, Statistical Division, WCCA, working under Col. Bendetsen. The Census War History draft chapter covering the work of the Vital Statistics Division, which was based on drafts prepared by that Division, indicate that Linder's mission was undertaken at the request of Dedrick "to assist Dr. Dedrick in studying the problems involved in a general identity registration for military purposes" [Dobbin, 1946b: 25].

Despite the relatively narrow focus implied by the title of Linder's report, its contents were far broader, covering a range of identity, administrative, and statistical functions that could be accomplished by different sorts of population registration systems. It is clear from both Linder's covering memorandum and his detailed report that whoever initiated the idea of population registration had a mixture of specific interests and only a vague awareness of realistic possibilities. (In his covering memorandum Linder observed, "as you know, the techniques and procedures for a registration can be planned only in general terms until a policy regarding the extent and the objectives of the system is more clearly formulated.") It also becomes clear, as one reads Linder's detailed report, that at least one military goal was to institute a universal civilian identity card [Linder, 1942: 4-5].

The introductory section to Linder's full report first stressed the "magnitude and complexity of the problem" of even "a general population identity registration" system, it then went on to reiterate the "necessity for a clear definition of purpose" [Linder, 1942: 1-3]. After examining and rejecting two alternatives Linder then concluded that

if military necessity or national policy requires that each person carry an identification certificate, then a new registration must be had for this purpose.

He then observed such a new registration and identity system would be of great value in other future registrations [1942: 4-5]. However, before turning to the body of his report, Linder continued [1942: 5]

Emphasis has already been given to the technical and administrative difficulties of general population registration for identity purposes. Mention should be made, also, of the political and psychological dangers. Traditional American thinking regarding freedom of action and thought might consider a mandatory identification register as an infringement of that liberty and the beginning of an American "gestapo." The political implications or effects of a compulsory identity registration might be considerable, unless a substantial part of the public clearly saw the necessity for it. Also the possibilities of "blacklist" inherent in an identification system are certain to arouse the opposition of labor groups.

The weight to be given such considerations by the War Department during time of war is a policy matter beyond the scope of this memorandum. However during peacetime, they have appeared to be effective barriers in the minds of population statisticians who have always been interested in creating a population register.

The balance of Linder's report is a detailed 30 page technical exposition with sections on "Types of registration systems," "Specific functions of registration systems," "Possible developments which might necessitate an identity registration," "The desirable functions of a registration system for each assumed condition," "A minimum identity registration system," "An optimum identity registration system," and "Additional technical problems to be considered."

Only two aspects this material seems pertinent to the present paper. First, it referred in several places to issues related to the statistical, demographic and analytical uses of data that might be generated from a population registration system [Linder, 1942: 6, 11,13, 25, and 30-33]. Second, it included among the 22 items listed in the final section for further consideration one surprising topic, "5. Value of loyalty oath" [1942: 34]. Although Linder's inclusion of material related to analytical uses of data seems a logical part of his report, the matter of the loyalty oath does not. We can only conclude that the issue arose because the idea of including a loyalty oath as part of the proposed general population registration system was suggested by someone who supported the system and was in a position to raise the matter with Linder.

Immediately upon receipt of Linder's report, Dedrick wrote to Bendetsen to give his own "findings and recommendations as an independent judgement of some of the factors involved in a general population registration" [Dedrick, 1942]. Not surprisingly given the letter he had drafted for Capt some months earlier [Capt, 1942a], Dedrick [1942] stressed the usefulness of a general registration of population to meet many military and civilian needs, including the provision of "quantitative population data on a current basis." He also considered that such a system was "administratively feasible" and could be accomplished at "relatively low cost." Emphasizing that the policy problems were of more immediate importance than the technical ones, Dedrick urged that the matter be approached so as allow for the possibility of establishing an integrated national system and avoiding uncoordinated efforts. "To accomplish this," he concluded, "will require consultation with other agencies in Washington and may require Presidential or Congressional approval."

In accordance with Dedrick's assessment, the locus of further work on the proposal for a national population registration system then shifted back to Washington, DC. For example, in early June, Professor Ogburn, perhaps by pre-arrangement wrote to Census Director Capt [Ogburn, 1942]

When I was in Washington recently, I heard several suggestions about the desirability of the U.S. Census Bureau keeping its central registration file for every individual in the United States. I wonder if this matter has ever been given any consideration by yourself or the other members and officers of the census [advisory] committee?

Capt [1942c] responded quickly by calling a meeting of the Census Advisory Committee for July 10-11 on the "general subject of national registration and its possible implications for the Census Bureau." Capt's letter, which was drafted by Halbert L. Dunn, Chief Statistician of the Bureau's Division of Vital Statistics, indicated the possibility that closely allied subjects such as delayed registration of births, the annual sample census, and population estimates might also be considered. Capt [1942d] also wrote to Professor Lowell J. Reed of Johns Hopkins, to invite him participate in the meeting to express his "point of view and thinking" as Chair of the Commission of Vital Records and indicating that

the purpose of the discussion will be to explore the need for and the feasibility of a general population registration, the urgency of such a program in the war effort, the interests of various agencies in special and general registrations, and the role of the Census Bureau in a population registration system.

Katherine Parker, not Dunn, drafted this letter. At that time Parker was in the Statistical Research Division, the Division Dedrick formerly headed.

Beginning with the July 1942 Census Advisory Committee meeting, a year-long battle then followed that turned on issues of statistical needs and policies, administrative needs and the public's response to those needs, bureaucratic rivalries, and human rights and privacy concerns. There were three primary sides to the ensuing controversies and, at least in nascent form, each was reflected in the July 1942 meeting of the Census Advisory Committee. The first position, reflected most clearly in the views of Halbert Dunn [CAC, July 1942: 1-2, and 4; US Census Bureau, 1943c; and CAC, January 1943: 10-11], was that the primary registration problem facing the country was the burden placed on the existing local and state registration offices by the thousands of native born Americans who were attempting to establish their citizenship by seeking a delayed certificate of birth. Proof of citizenship was required for most war-related work and since prior to 1935, most births went unregistered, large numbers of adults were filing applications for the delayed registration of their own birth.

The second position strongly and consistently advocated by Lowell J. Reed, both individually and as chairman of the Commission on Vital Records, was that population registration was very important for administrative and statistical purposes [CAC, July 1942: 3-5; Commission on Vital Records, 1942; Reed, 1942; and Commission on Vital Records, 1943: Forward by the Chairman]. In his zeal to advance the cause of general population registration, Reed often seemed to be hampered by a Commission that, like Dunn, gave priority to issues related to birth and death registration.

The third position, reflecting the views of senior Census Bureau management, was that general population registration simply represented an opportunity for the Census Bureau to both gain resources and "provide the nation with much needed current statistical information" [Capt, 1942f]. In advocating this position the Bureau was also hampered by the fact that those with primary over-all substantive responsibility for work in the area of population and vital statistics within the Bureau had at best quite mixed views on the advantages of general population registration (see, for example, Dunn [CAC, July 1942: 3], Linder [1942], and Truesdell [1943].

These conflicting positions were ultimately resolved in the Bureau of the Budget in the first few months of 1943. In the end, there was no general population registration system, the major urgently needed improvements in the system of vital registration were introduced, and the Census Bureau lost its responsibilities for vital statistics and related registration work [US Bureau of the Budget, 1943]. In reaching its decision the Bureau of the Budget was greatly assisted by Dunn, who was assigned to work in that agency to help resolve the issues on a half-time basis between September 1942 and March 1943.

Most of the details of these controversies are beyond the scope of this paper. Two aspects, however, seem particularly relevant and are therefore summarized here: (a) information on the possible historical roots of the proposal for population registration in the population statistics community and (b) the role of the non-technical policy issues raised in Forrest Linder's mission report.

If the origins of the idea of pursuing general population registration at the early stages of World War II had a civilian rather a military basis, as the evidence seems to indicate (see, for example, Capt [1942a] and Dobbin [1946b: 25]), how did the senior leadership of the Census Bureau decide the idea was worth pursuing? We offer two possibilities. First, as already indicated articles about the successful implementation of such comprehensive registration systems in Europe, particularly in the Netherlands, had recently appeared in American statistical and demographic journals. Second, one of the sessions at the December 1940 annual meeting of the American Statistical Association (ASA) was organized around the topic, "Problems of statistical control of the defense program." The lead paper in that session [Lang, 1941], by the Chief Statistician in the Office of the Assistant Secretary of War, cited in its opening paragraph the presidential address by the ASA president in 1917, Allyn A. Young. This address discussed analogous challenges for the statistical community occasioned by World War I [Young, 1918]. Decrying the lack of statistical preparedness he found when the country entered that war, Young speculated on "what might have been accomplished ... when we entered the war, or even when our participation in the war came to be a serious possibility" [1918: 878-879]. After first calling for an up-to-date population census, Young went on to suggest that

The population census, even, might well have been used as a starting point for continuous local population registers, under the control of the police or of local registration officials. In all this work it would be sufficient if the general tabulations showed classes only, but information respecting individuals ... ought to be available, by means of proper formalities, to the appropriate branches of the War administration. [1918: 879]

Given the active role that Dedrick and others in the Bureau played in both the ASA and the PAA, it seems highly likely that they were aware of both sets of papers. What is not clear is whether Young's comment about making "information respecting individuals ... available ... to the appropriate branches of the War administration," also had an impact.

Linder's report was one of the basic documents before the Census Advisory Committee meeting of July 1942 when the population registration proposal was first discussed [CAC, July 1942: 2] and his report framed many of the subsequent technical and policy discussions about the proposal. Two policy issues stand out from Linder's report (a) the need for clarity about objectives and (b) the potential privacy and human rights threats lurking in the population registration proposal. The first point was acknowledged by Dedrick [1942], stressed by Truesdell [CAC July 1942: 5], and became a major focus of the work of the Commission of Vital Records.

The privacy and human rights issues raised by Linder were ignored by Dedrick and Capt. Nevertheless, they were taken up, one way or another, by the other major participants in the registration proposal policy debates. For example, Dunn included the relationship between the proposed registration activities to the "police function and espionage in the minds of the public" among a list of dangers for the Bureau related to the proposal [CAC July 1942: 3]. On the other hand, the Commission on Vital Records [1942:11-12] and Reed [1942: 3-4] in arguing for the population registration proposal, considered it necessary to attempt to counter Linder's "American gestapo" comment. Finally, the report of the Bureau of Budget, which marked the formal end of the population registration proposal, referred to the need to avoid any system that would "lead to the compilation of a secret dossier on the American people" [US Bureau of the Budget, 1943: 6].

V. Long-term Consequences

We now turn to the utilitarian question of whether failures by a statistical agency to aggressively defend the rights of all respondents, even those perceived to be "enemies" or members of a despicable class, have any long-term implications for the agency's mission. Specifically, did the Census Bureau's actions in 1941 and 1942 with respect to Japanese Americans and its subsequent treatment of the issue have any consequences for population censuses in later decades? We can provide no clear-cut answer to this question, because neither the Bureau nor anyone else seems to have investigated the issue explicitly. Limited anecdotal evidence coupled with suggestive results from two quantitative studies make clear that it is a subject that should be investigated explicitly.

Although the Census Bureau has repeatedly denied that it betrayed its confidentiality pledges to respondents in the 1940 Census, citing a range of legal authorities for its position, the issue has never gone away in the press and in segments of the responding public. For example, at the eve of the 1980 Census director Barabba was quoted in the NY Times (3/31/80, B1) referring to the use of 1940 Census results to assist in the detention of Japanese Americans, "With all the hysteria of Pearl Harbor, we were not allowed to give up personal information." Less than a month later, the same paper (4/27/80, 62) reported the following exchange at a meeting of the Community Census Committee of Chinatown,

Ruby Schaar of the Japanese-American Citizens League said, "There have been a lot of rumors about the use of the census data." She questioned Government assurances that it would not be used against citizens or illegal aliens. "During World War II, printed reports were used to identify where Japanese people lived," Mr. Barabba responded. "But we never gave out any specific names and addresses."

A similar picture of people talking past each other - a mixture of carefully-phrased legal distinctions and inaccurate language on the one hand, and concerns over real harms arising from cooperation with the census, on the other hand - may be found in the sharp colloquy between Dr. Dedrick and Judge William M. Marutani, the only Japanese-American member of the Commission on Wartime Relocation and Internment of Civilians, at the conclusion of Dedrick's testimony before the Commission [Dedrick, 1981b: 187-190].

During the course of preparing this paper, the authors have also heard from some African Americans, who grew up on the West Coast, that their parents tried to avoid being enumerated in the census, citing the experience of the Japanese Americans and the 1940 Census. Whether this source of census omission had or has a measurable impact on the high over-all omission rate of African Americans in U.S. population censuses is not known. Others have commented on the same phenomenon among Japanese Americans and other Asian-Americans living on the West Coast [Kondo, 1999]

Census workers say that even some longtime Asian American residents [in Los Angeles county] distrust the government tally. Some Japanese Americans who are first or second generation still harbor fears that census numbers were used to identify them for placement in internment camps.... Convincing perennially reluctant respondents will not be easy.

"I talked to one older Filipino couple at a festival, and they had been living in the Valley for 50 years and never filled out the forms," said Susan Ng, a Census Bureau community specialist for the Asian community. "That's amazing - they missed about five censuses."

Fan [1999] reported similar concerns about the census among Asian Americans in central California.

Relevant published studies of empirical data on census and survey omissions tend to be so broadly focused that the results provide few answers, although a few underscore the need for new studies or at least a re-analysis of existing data. No study providing specific estimates of the coverage of Japanese Americans in a recent census was identified. Fein's study of racial and ethnic differences in census omissions based on estimated gross omission rates from the 1980 Census found over-all omission rates for Asian Americans (6.7 percent) substantially above that for non-Hispanic white (4.0 percent) [1990: 290]. Similarly, the omission rate for the total population in the Pacific census division (6.1) was higher than that encountered in six of the eight other census divisions [1990: 291]. No estimates of omission rates cross-classified by race/ethnicity and census division were presented. If the anecdotal evidence just discussed reflects a general pattern one would expect that such cross classifications would show particularly high omission rates for a number of individual non-white racial/ethnic categories in the Pacific division.

Although Singer, Mathiowetz, and Couper [1993: 479] concluded that confidentiality concerns had a significant but minor direct impact on nonresponse among whites and a more complex impact among African Americans, their findings were based on research into generic concerns held by broad population groups. Again the more salient research question to the problem at hand is to ask about concerns and behavior related the internment of the Japanese Americans and would specifically test hypotheses related to individual census racial/ethnic categories controlled for West Coast residence.

International survey research experience may also provide insights into the long-term persistence of mistrust and the degree to which this mistrust may effect response behavior both in specific communities and the population at large. Again the matter has never been explicitly studied, but one suggestive example is worth considering. One may recall the extensive and direct use of population data systems in the Netherlands in targeting Jews and gypsies for the Holocaust and that the death rates among Jews and gypsies was higher in the Netherlands than in any other occupied Western European country [Moore, 1997; Seltzer, 1998]. Is it a coincidence or is there a causal link between the Dutch experience with the adaptation of their population data systems to serve the Holocaust and the response problems that Statistics Netherlands has faced for decades? These problems include such massive nonresponse to the 1973 population census that the census had to be canceled. No subsequent general enumeration was ever attempted. Moreover, in a recent study of trends in survey response among 16 industrialized countries over the period 1983-1997, the Netherlands had the highest labor force survey nonresponse and refusal rates [de Heer, 1999: tables 1 and 3]. For example, in 1997 the refusal rate was 28 percent, with little variation over time. A similar pattern was found in expenditure surveys [1999, table 4]. By contrast, the corresponding rates for all other countries studied were far lower and many displayed clear secular trends. The experience of Statistics Netherlands is certainly consistent with the conclusion of Singer, Mathiowetz, and Couper [1993: 479] that "the crucial variable appears to be trust in the integrity of the data collection agency, not the nature of the assurance given to respondents."

VI. Discussion

(1) Why did this happen?

It is not easy to sort out the multiple motivations of those whose actions seem to actively assist human rights abuses. Moreover, such an effort at understanding motivation is in many senses irrelevant to those who suffered from the abuse. Nevertheless, some examination of motivation can be helpful in preventing future abuses. In the present case, such a discussion may contribute to the heightened awareness by all those working on the collection and use of population statistics that various of their well-intentioned actions and inactions can have human rights-related consequences. This is an important matter for all of us to be aware of since our field and the institutions in which most of work are generally considered to be far-removed from such concerns.

Statisticians and demographers, and others working in and leading agencies responsible for major population statistics programs are motivated by numerous personal, institutional, and societal factors. For example, as the record reviewed in this papers shows, we can the following motivating factors seemed to be at work in the Bureau's contributions to the targeting and the interning of Japanese Americans: (a) importance of serving military needs in wartime, (b) it was the Director's decision, (c) the need for the Bureau to establish its continued relevance in time of war, (d) opportunity to develop and test new methods, (e) expanding the use Census Bureau outputs, (f) we followed the law, (g) concerns over the level of funding for the Bureau's programs, and (h) dedication to the ideal of the Bureau's objectively providing data to all users without regard to the nature of the use. These motivations can be grouped under three broad headings: patriotism, professionalism, and institutional loyalty.

During the period in the late 1930's and early 1940's when the United States was moving from defense preparations to actual war, these of three broad factors interacted in changing and complex ways. Through 1941, as the likelihood of war increased, the data from the new registration systems poured in. Meanwhile the processing of the other major population collection, the 1940 Census, slowed. Budgets were cut for the census, and requests for data from national defense grew. (Anderson, 1988). War was clearly on the horizon, and the officials knew it. It is hard not to feel the sense of impending war in the routine memos and documents of the agencies, as the officials expressed fear of the coming threat, disappointment at the likelihood that the ambitious publication and evaluation plans for the 1940 Census might have to be scrapped, and worry whether the nation was up to the task of defending itself. The emerging results of the 1940 Census were not completely satisfying in that regard either. The decline in the population growth rate, the damage to the human capital of the nation, were all too evident in the reports being released in 1941.

Nevertheless, to this point, the basic tone of Bureau officials was one of respect for the rights of individuals on whom data was to be collected, and concern for need to maintain standards of privacy and confidentiality of individual records. Old hands remembered that selective service officials had asked for, and received, information on individuals from the 1910 Census for the prosecution of draft evaders. William Mott Steuart, Census Director during the 1920s, and then serving on the American Statistical Association-American Economic Association Census Advisory Committee (CAC) was aware of these incidents. At a meeting in the fall of 1940, he asked the current Director, William Lane Austin, to report if the Justice Department had requested names of individual aliens from the bureau, or requests from the Army or the Navy for names of individual manufacturers. Austin reported there had been no such requests [CAC, October 1940].

The concern with confidentiality, the protection of individual information, and respect for the sensitivity of the data did not survive the outbreak of war in tact. The surprise attack on Pearl Harbor on December 7, 1941 transformed the relationship between the statistical system and the other agencies of the federal government. The declaration of war against Japan on December 8, and on December 11 against Germany and Italy, placed fundamental new challenges on the population data systems of the nation. At a stroke, it turned roughly one million aliens from Germany, Japan and Italy into 'enemy aliens,' liable under international law to increased scrutiny, surveillance and possible internment. And after two years of the coverage of war in Europe and Asia, Congressional debate and media coverage about the loyalty of aliens, the devastation of Pearl Harbor inspired fear in many quarters that enemy agents within the country had prepared Germany submarines or Japanese air strikes on additional domestic targets. At this point, all the operative forces within the Bureau appeared to be aligned toward the use of Census Bureau data and expertise against the Japanese Americans on the West Coast. In any event, we could not find a record of even a single voice within the Bureau or its advisory committee raising a doubt about the actions taken. (The Bureau was aware of doubts by at least one external source within the profession: Dedrick's "sociological friends at Berkeley.")

Many of the same factors that led the Bureau to actively assist in the internment effort also seemed to be at work in its support for the proposal for a general population registration system. Only in the case of the registration proposal, one Census Bureau statistician (Linder) raised a set of policy issues that clearly added to the baggage that the advocates of the proposal had to deal with. Linder, while demonstrating the same broad motivations of patriotism, professionalism, and institutional loyalty as did his colleagues when they dealt with internment, defined these concepts more broadly. As we have seen, relatively quickly other voices reiterated his points about the population registration proposal and raised other questions about it. Nevertheless, his initiative and courage also remains an example of the power of a single voice to raise an issue and effect an outcome.

(2) Critical decision points

The pressure of total war usually puts great strains on normal decision-making processes. In such fast-moving circumstances, there are also often opportunities for changes in direction. We now turn to an examination of a series of critical decisions involving both the Census Bureau's involvement with the internment program and the early stages of the effort to establish a general population registration system for military and statistical purposes. This examination illustrates, with respect to the Bureau's involvement in the internment issue, the many opportunities that existed for the outcome to have turned out quite differently. It is also useful to contrast the history of the two events discussed in this paper - the Census Bureau's involvement with the internment of the Japanese Americans and its aftermath and the Bureau's involvement with the proposal for population registration - in terms of a review of critical decisions.

(a) Japanese-American internment

$ Monday morning, December 8, 1941 the decision to tabulate the "Japanese" using the item on race rather than the item on country of birth, as was done for the Germans and Italians two days later.

$ Balance of December 1941, the failure to reconsider this decision.

$ January 10-11, 1942, the failure of anyone present at the Census Advisory Committee meeting to question the use of small area census data to target Japanese Americans based on their 1940 Census responses or to question the Director's willingness to provide names and addresses to the military, if requested.

$ February 1942, the decision by the Census Bureau to send Dedrick, a senior division chief, to the West Coast as Bureau staff member to assist an operational program (i.e., the Army's Wartime Civil Control Administration) and use 1940 Census results to target Japanese American members of the responding public for forced migration and internment.

$ February 1942, the related decision by the Census Bureau to provide Dedrick and the WCCA with mesodata and maps from the 1940 Census down to the block and enumeration district level.

$ January-March 1942, the decision to base the internment on the1940 Census concept rather than using the alien registration lists corrected for omissions.

$ 1943, the decision by Dedrick to obfuscate in the Army's final report of the internment, how information from the 1940 Census was used.

$ 1946, the decision by Dedrick to continue the obfuscation in his contributions to the Census War History project.

$ 1975, the decision by the Census history staff to do no independent research to respond to a query from Roger Daniels, leading to the first story, attributed to Dedrick, of the Bureau's effort to protect of Japanese Americans.

$ 1980-1982, the failure the Census history staff, successive Census directors, and the larger population statistics community to adequately re-examine the historical record in light of repeated questions by members of the Japanese American community and others, including the US Commission on Wartime Relocation and Internment of Civilians, about the Bureau's involvement in the internment effort.

$ 1991, the decision by the Census history staff again to do no new research in writing about the Bureau's historical record on privacy and confidentiality issues.

$ 1995, the decision by former Census Director Bryant to acknowledge that however valid the Bureau's legal position may have been, the Japanese American community had justifiable reasons for believing that the spirit of the confidentiality assurances that accompanied the 1940 Census had been violated.

$ 1999, the decision of the Census Bureau to reuse the old 1980 story about its defense of Japanese American interests at the time of the internment in some of the training materials used in the 2000 Census.

We are pleased to note that in the current year (i.e., 2000) the Census Bureau has made several decisions that seem to indicate that Bureau is ready to revisit the past and learn from it. They have already attempted to correct the mistaken information that appeared in the census training materials.


(b) The population registration proposal

$ February 1942, the decision by Capt and Dedrick to begin pursuing the idea of general population registration.

$ Spring 1942, the decision by Dedrick and Bendetsen to arrange for Forrest Linder's mission to the West Coast to examine population registration.

$ May 1942, the decision by Linder in his initial mission report to address policy issues in addition to technical issues.

$ July 1942, the decision by Dunn to again raise Linder's policy concerns in presenting the proposal for the population registration system to the Census Advisory Committee and the decision by Committee after discussing the matter that further consideration was necessary.

Although a number of important decisions remained, including the ultimate one by the Bureau of the Budget in June 1943 to formally abandon the proposal, it appears that the die had already been cast once the rush to action had been delayed by the series of policy and technical questions raised by Linder, Dunn and Truesdell.

(3) Some implications of our findings for statistical policy and research

Although the past is interesting in its own right, it also has implications for the present and the future. Accordingly, we have drawn two sets of conclusions from the work done so far. The first set relates to statistical policy and the second set to needed further research. At this point, these conclusions and recommendations are set out with little or no argumentation since we consider they flow naturally from the findings presented earlier in the paper.

(a) Statistical policy

$ Strengthen the historical staff so that they have both the mandate and capability to move beyond simply documenting the census and survey procedural histories.

$ Observe the same degree of forthrightness with regard to policy errors that major statistical agencies, such as the Census Bureau, have traditionally practiced with regard to statistical errors.

$ Continue to strengthen technical protections against inadvertent and intentional response disclosure, including the use of mesodata to target out-of-favor and vulnerable populations.

$ Develop new legal safeguards that will protect against the abrogation of the confidentiality provisions of Title 13, such those enacted through the War Powers Act of 1942.

$ Drawing on such normative documents as the United Nations Fundamental Principles of Official Statistics (which may be found at www.cbs.nl/isi/fundamental.htm), the International Statistical Institute's Declaration of Professional Ethics (which may be found at www.cbs.nl/isi/ethics.htm), and the American Statistical Association's Ethical Guidelines for Statistical Practice, strengthen training that promotes awareness of ethical issues and thinking among the staff and leadership of federal statistical agencies.

$ Until strong technical and legal safeguards are developed and are in place, do not collect characteristics that define and identify "potentially censurable or vulnerable" groups [Begeer, de Vries, and Dukker,1986] on a full-count basis. In other words, the census item on "race," and certainly information on detailed racial categories, should no longer be collected or stored on a 100 percent basis. We recognize that, in the present circumstances, some may be tempted to reject this recommendation out of hand. We note, however, that our recommendation is far more limited than the position taken by the American Anthropological Association(AAA) on this subject. The AAA has recommended that the item on race be totally dropped from the census [Anderson and Fienberg, 1999: 173-174]. Limiting of collection of data on race, particularly involving detailed racial/ethnic categories, to samples will reduce substantially the possible misuse of such data, although it will not completely eliminate the all possible types of misuse. The approach we recommend would also enable the continued availability of a considerable body of data cross-classified by race, particularly at higher levels of geographic aggregation. This approach represents, in our view, a reasonable balance between the needs of many users for data classified by race/ethnicity and a prudent concern for public safety, given the failures of the past in the United States and elsewhere.

(b) Further research

As we indicated at the outset, our present paper is an interim report of a larger ongoing study of how population data systems were used to assist in the military defense of the United States between 1940 and 1943 and the human rights and statistical policy implications of this involvement. A related issue, also explored, was how those working with these systems, whether as technicians, managers, or members of advisory committees, used the national emergency to assist in the further development of the data systems themselves. Given the findings to date and our interim listing of statistical policy conclusions, several quite different areas for further research seem important. We have grouped the under three broad headings: (1) statistical research, (2) legal research, and (3) historical research. However, the final category could also have been called human rights research or statistical policy research.

(1) Statistical research
Research in two specific areas is suggested:

$ First, research into nonresponse and refusals aimed at improving our "understanding of nonresponse" [de Leeuw, 1999: 127], particularly in circumstances where population subgroups may perceive that harm was associated with past response. In the United States such research would include (a) a reanalysis of existing census and survey evaluation data to explicitly test the hypothesis of a relationship between internment and response behavior and (b) explicitly asking Pacific Coast populations about their response behavior and the possible impact of internment on such behavior.

$ Second, studies of the accuracy and precision of alternative methods of developing census coverage estimates if race/ethnicity were not available on a complete count basis. The absence of the race/ethnicity item on a complete count basis would have an impact on the present method of estimating the differential undercount by race, although it would not preclude estimates of the differential. It would be important to understand what impact the potential loss of complete-count race data might have on the quality of undercount estimates for the total population and for individual race/ethnicity groups and for different types of geographic aggregates. On the basis of this information alternative approaches to coverage estimation could be explored with a view to minimizing this loss.

(2) Legal research
Again research in two areas is suggested:

$ Both Dedrick in his testimony before US Commission on Wartime Relocation and Internment of Civilians [Dedrick, 1981b] and Census Director Barabba in a letter to Okamura [1980] referred to the possible impact on census confidentiality posed by the Second War Powers Act of 1942. Although Dedrick stated that he was not aware that the act was invoked in connection with individual data provided in the population census, that issue plus the related one of whether the possibility of such a use figured in the legislative history of the Act or the development of related Executive Order No. 9157 of May 13, 1942, that provided regulations for making disclosures of individual census schedules as permitted by the 1942 War Powers Act [US Census Bureau, 1942a: 29-30], could usefully be explored (see Seltzer [1999b]).

$ Research into legal mechanisms to protect against future legislative waiver's of the confidentiality protections of Title 13 should be undertaken.

(3) Historical research

What we label here as historical research, but could equally well be termed statistical policy or human rights research, is really the continuation of the present research that led to this paper. At a number of junctures in the paper we have indicated the specific research questions to be addressed. At this point, we would simply summarize the next steps in terms of data sources to be examined.

$ Further federal archival research covering: Record Group 29 (Census Bureau); the War Department, including the Office of the Assistant Secretary of War, the Office of the Provost General, and Western Defense Command; the intelligence agencies, both military and civilian; the WCCA and the WRA, particularly the statistical units of both organizations; the Office of the Solicitor, Department of Commerce; and the Commission on Vital Records.

$ Other archives: State of California; Earl Warren Oral History project and other holdings at Bancroft Library, University of California, Berkeley; and the collections of the personal and professional papers of key individuals.

$ California newspaper research: the set of California newspapers papers covered by Grodzins in his study of the role of the California press in the internment [Grodzins, 1949].
We would submit that further research along these lines is important for two reasons. First, to be clear about how the tools of our profession and how some of our former colleagues assisted in an acknowledged human right abuse. As former President Bush wrote to each recipient of redress payments

A monetary sum and words alone cannot restore lost years or erase painful memories; neither can they fully convey our Nation's resolve to rectify injustice and to uphold the rights of individuals. We can never fully right the wrongs of the past. But we can take a clear stand for justice and recognize that serious injustices were done to Japanese Americans during World War II. [Quoted by Tetsuden Kashima in his new forward to CWRIC, 1997].

Second, as one concerned statistical administrator has observed, "it is important to let light shine on the past so as to avoid possible mistakes in the future."

NOTES

The authors would like to acknowledge the research assistance provided by Ms Chieko Maene, a graduate student at the University of Wisconsin-Milwaukee. We would also like to thank Jane Berger, Roger Daniels, Steve Fienberg, John Macisco, Mary G. Powers, Herb Spirer, and Maurice Van Arsdol for research leads and suggestions provided at various stages of this work. Finally, we would like to thank Kenneth Prewitt and Paula J. Schneider of the Census Bureau for their encouragement to continue in the research underlying this paper. Although we are deeply grateful for the help received, it should be emphasized that the authors alone are responsible for the contents of the paper.

  • While not discounting the value of mesodata, the specific numeric example cited by Director Capt seems absurd. No one should attempt to expend any investigative resources to resolve a discrepancy of one person out of 800 in a specific community between an April 1940 census count and a January 1942 count or registration, given the possibility of births, deaths and migration during the intervening period or coverage or classification discrepancies between the two figures. Capt was an experience federal administrator who led the Bureau in one of its most creative periods. He himself, however, was not a statistician or a quantitatively oriented social scientist.
  • The issue has sometimes become more complex with respect to copies of census returns held by corporations, but the protection afforded the census information provided by individual persons has not been seriously questioned by any court after the modern protections were codified with the adoption of the Census Act of 1929. Over the decades, these legal protections have been strengthened.
  • In fact Schmid appears to have been opposed to internment himself, even though he worked for the WCCA for several months during 1942 before returning to Seattle. According to Van Arsdol and Wendling [1995: 13], Schmid befriended interned colleagues and worked to limit the geographic scope of one of the exclusion zones. In his study Social Trends in Seattle [Schmid, 1944] he certainly presented a positive picture of the excluded population even as he demonstrated the same graphic and cartographic skills in data presentation that can also be found in many of the reports of the WCCA and the Western Defense Command's Final Report [US Army, 1943].
  • For example, the WRA's Headquarters' Statistics Section consisted of four units: Operating Statistics, Analysis, Master File, and Evacuee Records [US War Relocation Authority, [1946a:16]. Equally suggestive are the two, and only two, subentries under the term "statistics" in the detailed subject-matter index to the WRA's Administrative Manual. These subentries are: "Individual Evacuee File" and "Individual File, Contents of" [US War Relocation Authority, 1945: xix].
  • Dedrick's misstatement about the purpose of Linder's 1942 mission to California was almost certainly deliberate. Later the same year, the Census Director had immediately agreed [Capt, 1942e] to the Army's request [Bendetsen, 1942b] that "no publicity should be given which would in any way indicate that we are or were at any time considering general registration."
  • The illustrations on pages 7 and 8 this Bureau publication are remarkably similar to some of those put out in the 1930's by the German affiliate of the Hollerith company and reproduced in Aly and Roth [1984]. The parallelism here we believe is not due to any direct connection, but to a common awe in the power of a new technology to assist in the tasks of the state.
  • By September 1947, Dedrick had been named as Coordinator, International Statistics for the Bureau. He served in this capacity for several decades making outstanding contributions in the field of technical cooperation and training in the field of statistics.
  • The language used here is remarkably similar to that used by Dedrick in 1946, and quoted in section I.D. above, to describe the War Department's request for his services.
  • The apparent contradiction between Dedrick's 1975 and 1981 statements may be explained in several different ways: for example, (a) by the slightly different terminology used in each statement (i.e., Census Bureau vs. Commerce Department), (b) because Bohme seriously misquoted him in 1975, or (c) by Dedrick's use of the Census War History project materials to refresh his recollection prior to his 1981 appearance before the Commission on Wartime Relocation and Internment of Civilians.
  • It appears that Frederick Bohme is also largely responsible for this account [US Census Bureau, 1980: xviii].
  • Implicit in this approach is the assumption that those who made the selection of materials used in the existing archival compilations were neither aware of or particularly interested in data issues, since these issues would have been overwhelmed by other more pressing priorities for inclusion.
  • Prewitt [1987: 269-270] made much the same point when he observed that "for two centuries we have had a statistical practice of racial classification which undoubtedly has contributed the continuing salience of race in American society."
  • Census data provided an aggregate check and comparison with the alien registration data. The 1940 Census included questions on place of birth and citizenship status. Tabulations for the alien foreign born (reported in 1943) also reported many fewer aliens than did the 1940 registration. The census reported about 3.5 million aliens and 835,000 foreign born residents with no citizenship status reported [U.S. Census Bureau, 1943b:10].
  • In 1942 Dr. Linder was Assistant Chief Statistician for Vital Statistics at the Bureau. His mission to San Francisco lasted from May 1 to 10. Subsequently, he joined the U.S. Navy on military leave and assisted in the establishment of the Medical Statistics Branch in the Office of the Surgeon General, U.S. Navy [Dobbin, 1946b: 23 and 25]. After the end of the War, he created and led (a) the population and vital statistics program in the United Nations Statistics Division, (b) U.S. National Center for Health Statistics, and (c) the international PopLab program of the University of North Carolina at Chapel Hill.
  • During his mission Linder was reported to have discussed the general subject of registration "with officials of the Western Defense Command and with the Assistant Superintendent of the California State Department of Education, who had been requested ... to develop plans for a general registration" [Dobbin, 1946b: 26]. Subsequently, in February 1943, an effort to introduce a new registration form containing a "loyalty oath" among the interned Japanese-Americans was a major source of discontent for many among this detained and loyal population [CWIRC, 1982: 191-197; Weglyn, 1996: 134-155].
  • Estimates of the net undercount have been prepared by the Census Bureau for the 1990 Census by race and Hispanic origin for each state based on the unadjusted and Public Law 94-171 adjusted census data and released on the Bureau's web site (see www.census.gov/dmd ). These estimates for the 1990 Census indicate that net omissions of Asian and Pacific Islanders in California and Washington (2.2 percent) were at about the same level as this group nationally (2.3 percent). Further research is required to determine whether or not this represents a change from the experience of the previous censuses.
  • On a closely related subject, Congress has barred items on religion from all population-based federal censuses and surveys.
  • Although we made extensive use of the material in particular parts of Record Group 29, in the time available most of Record Group 29 could not be examined.

References

Aly, Götz and Karl Heinz Roth. 1984. Die restlose Erfassung: Volkszählen, Identifizieren, Aussondern im Nationalsozialismus [Complete Registration: Population Census, Identity, and Separation under National Socialism]. Berlin: Rotbuch Verlag.
Anderson, Margo J. 1988. The American Census: A Social History. New Haven (CT): Yale University Press.
Anderson, Margo and Stephen E. Fienberg. 1999. Who Counts? The Politics of Census Taking in Contemporary America. New York: Russell Sage Foundation.
Austin, W.L. 1940. "An official view of proposed questions for 1940 Census." Letter to the editor dated February 27, 1940. The New York Times, March 3, 1940, Section N, page 8.
Barabba, Vince. 1980. Letter to Raymond Okamura. December 16, 1980. Quoted in part in Okamura [1981].
Begeer, W., W.F.M. de Vries, and H.D. Dukker. 1986. "Statistics and Administration." Eurostat News, Special Edition - Protection of privacy, automatic data processing, and progress in documentation.
Bendetsen, Karl R. 1942a. "Location of certain persons of Japanese ancestry." Memorandum to Dr. Dedrick. July 3, 1942. CWRIC microfilm.
. 1942b. Letter to J.C. Capt, July 4, 1942. Papers of Philip Hauser (item 146); US Census Bureau, Record Group 29; National Archives Building, Washington, DC.
Bohme, Frederick G. 1975. Letter to Roger Daniels, dated December 3, 1975. Records relating to the War History Project (item 156); US Census Bureau, Record Group 29; National Archives Building, Washington, DC.
Bohme, Frederick G. and David N. Pemberton. 1991. "Privacy and confidentiality in the US Censuses -- A history." Paper prepared for presentation at the annual meeting of the American Statistical Association, August 18-22, Atlanta, GA.
Bryant, Barbara Everitt and William Dunn. 1995. Moving Power and Money: The Politics of Census Taking. Ithaca (NY): New Strategist Publications
Capt, J.C. 1942a. Letter to Dr. William F. Ogburn, February 26, 1942. General Records of the Census Advisory Committee (item 148); US Census Bureau, Record Group 29; National Archives Building, Washington, DC.
. 1942b. Memorandum to [John] McCloy, dated March 18, 1942. CWRIC microfilm from Record Group 210, Commerce Department, Chrono file.
. 1942c. Letter to Dr. W. F. Ogburn, June 18, 1942. General Records of the Census Advisory Committee (item 148); US Census Bureau, Record Group 29; National Archives Building, Washington, DC.
. 1942d. Letter to Dr. Lowell J. Reed, July 1, 1942. General Records of the Census Advisory Committee (item 148); US Census Bureau, Record Group 29; National Archives Building, Washington, DC.
. 1942e. Letter to Colonel Karl R Bendetsen, July 13, 1942. Papers of Philip Hauser (item 146); US Census Bureau, Record Group 29; National Archives Building, Washington, DC.
. 1942f. Letter to Lowell J. Reed, October 10, 1942. Exhibits at the Census Advisory Committee meeting October 30-31, 1942; Minutes of Meetings, Correspondence and Reports, July 1942 to August 1943; General Records of the Census Advisory Committee (item 148); US Census Bureau, Record Group 29; National Archives Building, Washington, DC.
. no date. Memorandum from the Director to All Division Chiefs, "Preparation for the War History Project." Undated copy. Records relating to the War History Project (item 156); US Census Bureau, Record Group 29; National Archives Building, Washington, DC.
Census Advisory Committee (CAC). October 1940. Minutes of meeting of October 11-12. Census Advisory Committee Meeting, October 11 and 12, 1940; General Records of the Census Advisory Committee (item 148); US Census Bureau, Record Group 29; National Archives Building, Washington, DC.
. January 1942. Full transcript of meeting of January 9-10. Census Advisory Committee Meeting, January 9 and 10, 1942; General Records of the Census Advisory Committee (item 148); US Census Bureau, Record Group 29; National Archives Building, Washington, DC.
. July 1942. Summary report of the meeting of July 10-11. Census Advisory Committee Meeting, July 10 and 11, 1942; General Records of the Census Advisory Committee (item 148); US Census Bureau, Record Group 29; National Archives Building, Washington, DC.
. October 1942. Summary report of the meeting of October 30-31. Census Advisory Committee Meeting, October 30 and 31, 1942; General Records of the Census Advisory Committee (item 148); US Census Bureau, Record Group 29; National Archives Building, Washington, DC.
. January 1943. Summary report of the meeting of January 22-23. Census Advisory Committee Meeting, January 22 and 23, 1943; General Records of the Census Advisory Committee (item 148); US Census Bureau, Record Group 29; National Archives Building, Washington, DC.
. April 1943. Summary report of the meeting of April 23-24. Census Advisory Committee Meeting, April 23 and 24, 1943; General Records of the Census Advisory Committee (item 148); US Census Bureau, Record Group 29; National Archives Building, Washington, DC.
Chapman, Bruce. 1982. Letter to Raymond Okamura, dated October 21, 1982. CWRIC microfilm.
Clark, Tom C. 1972. Interview for the Earl Warren Oral History Project. The Bancroft Library, University of California, Berkeley.
Commission on Vital Records, 1942. "Report of The Commission on Vital Records, October 12, 1942." Exhibits used with the Advisory Committee Meeting, January 22-23 1943; General Records of the Census Advisory Committee (item 148); US Census Bureau, Record Group 29; National Archives Building, Washington, DC.
. 1943. "Report of The Commission on Vital Records." Final Meeting of the Committee - October 12, 1942. Released for distribution 2/3/43 with a new "Forward by the Chairman." Vital Records Commission; Papers of Philip Hauser (item 146); US Census Bureau, Record Group 29; National Archives Building, Washington, DC.
Conn, Stetson, Rose C. Engelman, and Byron Fairchild. 1964. The Western Hemisphere: Guarding the United States and its Outposts. Office of the Chief of Military History, Department of the Army, in the series United States Army in World War II. Washington, D.C.: US Government Printing Office. [Chapter V "Japanese Evacuation from the West Coast," pp. 115-149.]
Corcoran, Thomas F. 1963. "On the confidential status of census reports." The American Statistician 17 (3): 33-40.
Daniels, Roger. 1981. Concentration Camps: North America. Japanese in the United States and Canada during World War II. Malabar (FL): Robert E. Krieger. [Revised edition].
_____ . 1982. "The Bureau of the Census and relocation of the Japanese Americans: A note and a document." Amerasia Journal 9(1): 101-105.
_____ (ed.). 1989. American Concentration Camps. A Documentary History of the Relocation and Incarceration of Japanese Americans, 1942- 1945. 9 volumes. New York: Garland Publishing.
. 1997. "L'Internamento di "Alien Enemies" negli Stati Uniti durantela seconda guerra modiale." [Internment of 'Alien Enemies' by the United States during World War II]. In Acoma: Revista Internazionale di Studi Nordamericani. Rome 11: Estate autunno, pp.39-49.
Dedrick, Calvet L. 1940a. Memorandum to Dr. Truesdell concerning alien registration. June 28, 1940. Population memorandums; Population, International statistics; General records maintained by Calvert Dedrick, 1935-42 (item 210); US Census Bureau, Record Group 29; National Archives Building, Washington, DC.
_____. 1940b. Memorandum to Dr. Truesdell concerning alien registration. July 29, 1940. Population memorandums; Population, International statistics; General records maintained by Calvert Dedrick, 1935-42 (item 210); US Census Bureau, Record Group 29; National Archives Building, Washington, DC.
_____. 1942. Memorandum to Colonel Karl R. Bendetsen, May 9, 1942. "Plans for a general registration of the civilian population." Papers of Philip Hauser (item 146); US Census Bureau, Record Group 29; National Archives Building, Washington, DC.
_____. 1946. Memorandum to [John V.] Dobbin. "Activities of the Bureau of the Census relating to the War Department evacuation of Japanese from the west coast," July 1. Records relating to the War History Project (item 156); US Census Bureau, Record Group 29; National Archives Building, Washington, DC.
_____. 1981a. Opening Statement of Calvert L. Dedrick to CWRIC, November 2, 1981. Photocopy of statement.
_____. 1981b. Testimony before CWRIC, November 3, 1981, pp. 170-190. Photocopy of transcript.
Dedrick, Calvert and Morris Hansen. 1937. The Enumerative Check Census: Census of Unemployment, 1937. Vol .4. Washington, D.C.: US Government Printing Office.
de Heer, Wim. 1999. "International response trends: Results of an international survey." Journal of Official Statistics 15(2): 129-142.
de Klerk, F.W. 1998. The Last Trek - A New Beginning. The Autobiography. London: Macmillan.
de Leeuw, Edith G. 1999. "Preface." Special Issue on Survey Nonresponse. Journal of Official Statistics 15(2): 127-128.
Dobbin, John V. 1946a. "How the Census Bureau Helped the Program to Evacuate Japanese from West Coast Cities." Marked typescript draft and related covering memorandum from Dobbin to Dedrick, dated July 16, 1946. Records relating to the War History Project (item 156); US Census Bureau, Record Group 29; National Archives Building, Washington, DC.
. 1946b. "Vital Statistics - War History." Typescript of draft chapter initially prepared by the Vital Statistics Division. Records relating to the War History Project (item 156); US Census Bureau, Record Group 29; National Archives Building, Washington, DC.
Duncan, Joseph and William Shelton. 1978. Revolution in United States Government Statistics. Washington, D.C.: US Government Printing Office.
Fan, Maureen. 1999. "Immigrants shy of census: California stands to gain financially from accurate count." San Jose Mercury News. 6/12/99
Fein, David J. 1990. "Racial and ethnic differences in U.S. Census omission rates." Demography 27(2): 285-302.
Grodzins, Morton. 1949. Americans Betrayed: Politics and the Japanese Evacuation. Chicago: University of Chicago.
Hutchinson, Edward P. 1981. Legislative History of American Immigration Policy, 1798-1965. Philadelphia: University of Pennsylvania Press.
Ibata, David. 2000. "1990 Census Undercount Cost City, Suburbs Plenty of Money Study Says."Chicago Tribune, Internet Edition. January 30, 2000.
Irons, Peter. 1983. Justice at War. New York: Oxford University Press.
Kondo, Annette. 1999. "Accurate Count Sought for Asian Immigrants." Los Angeles Times. July 5, 1999.
Lang, Richard O. 1941. "Problems of statistical control -- Military aspects." Journal of the American Statistical Association 36(213): 11-17.
Linder, Forrest E. 1942. "Memorandum relating to a general population identity registration for military purposes" and covering letter to Calvert L. Dedrick, both dated May 9, 1942. Papers of Philip Hauser (item 146); US Census Bureau, Record Group 29; National Archives Building, Washington, DC.
Lorimer, Frank, Ellen Winston and Louise Kiser. 1940. Foundations of American Population Policy. New York: Harper and Brothers.
Methorst, H.W. 1936. "The New System of Population Accounting in the Netherlands." Journal of the American Statistical Association 31(196): 719-722.
________ 1938. "The New System of Population Accounting in the Netherlands." Journal of the American Statistical Association 33(204): 713-714.
Moore, Bob. 1997. Victims and Survivors: The Nazi Persecution of the Jews in the Netherlands, 1940-1945. London: Arnold.
Neuman, Gerald L. 1996. Strangers to the Constitution: Immigrants Borders and Fundamental Law. Princeton: Princeton University Press.
Ogburn, William F. 1942. Letter to J.C. Capt, June 9, 1942. General Records of the Census Advisory Committee (item 148); US Census Bureau, Record Group 29; National Archives Building, Washington, DC.
Okamura, Raymond Y. 1981. "The myth of census confidentiality." Amerasia Journal 8(2): 111-120.
Prewitt, Kenneth. 1987. "Public Statistics and Democratic Politics." In The Politics of Numbers, William Alomso and Paul Starr (eds.). New York: The Russell Sage Foundation.
Reed, Lowell J. [?]. 1942. "Explanation of a National Population Registration System Adaptation for War Purposes of Plan IV of the Commission of Vital Records." Unattributed position paper dated November 12, 1942 and marked "confidential." Papers of Philip Hauser (item 146); US Census Bureau, Record Group 29; National Archives Building, Washington, DC.
Rémond, René, et al. 1996. Le 'Fichier Juif' : Rapport de la Commision présidée par René Rémond au Premier Ministre. Paris: Plon éditeur.

Schmid, Calvin F. 1944. Social Trends in Seattle. University of Washington Publications in the Social Sciences, Volume 14. Seattle: University of Washington Press.
Seltzer, William. 1998. "Population statistics, the Holocaust, and the Nuremberg Trials." Population and Development Review 24(3): 511-552.
______ . 1999a. "Excluding Indians Not Taxed: Federal Censuses and Native-American in the 19th Century." Paper presented at the 1999 Joint Statistical Meetings, Baltimore, MD, August 1999.
______ . 1999b. "Second War Powers Act of 1942: Background and Some Needed Research." Memorandum dated December 28, 1999.
Singer, Eleanor, Nancy A. Mathiowetz, and Mick P. Couper. 1993. "The impact of privacy and confidentiality concerns on survey participation: The case of the 1990 U.S. Census." Public Opinion Quarterly 57(4): 465-482.
Søbye, Espen. 1998. "Et moerkt kapittel i statistikkens historie?" [A dark chapter in the history of statistics?]. Samfunsspeilet 8, no.4 (10 September). Oslo: Statistics Norway.
tenBroek, Jacobus, Edward N. Barnhart, and Floyd W. Matson. 1954. Prejudice, War and the Constitution: Japanese American Evacuation and Resettlement. Berkeley: University of California Press.
Thomas, Dorothy S., 1937. Review of De Bevolkingsboekhouding by J.L. Lentz. American Sociological Review 2(1): 117-118.
Toland, John. 1982. Infamy: Pearl Harbor and its Aftermath. Garden City (NY): Doubleday.
Truesdell, Leon E. 1941. "New features of the 1940 Population Census." Journal of the American Statistical Association 36(215): 361-368.
. 1942. "Functions of Population Registration." Notes prepared for the July 1942 Census Advisory Committee Meeting, July 9, 1942. General Records of the Census Advisory Committee (item 148); US Census Bureau, Record Group 29; National Archives Building, Washington, DC.
______ . 1965. The Development of Punch Card Tabulation in the Bureau of the Census, 1890-1940, with Outlines of Actual Tabulation Programs. US Department of Commerce, Bureau of the Census. Washington, D.C.: US Government Printing Office.
US Army. Western Defense Command and Fourth Army. 1943. Final Report: Japanese Evacuation from the West Coast, 1942. Washington, D.C.: US Government Printing Office.
US Bureau of the Budget. 1942. "Priority classification of executive departments and agencies, required by Executive Order No. 9243. Circular No. 400" Memorandum to heads of executive departments, independent establishments and agencies, dated September 25, 1942. National Defense; Papers of Philip Hauser (item 146); US Census Bureau, Record Group 29; National Archives Building, Washington, DC.
. 1943. "Measures Relating to Vital Records and Vital Statistics." Message of the President of the United States Transmitting the Report of the Bureau of the Budget. (House Document No. 242, 78th Congress, 1st Session, June 24, 1943) Washington, DC: US Government Printing Office.
US Census Bureau. 1941a. Japanese Population of the United States and its Territories and Possessions. Sixteenth Census of The United States. Series P-3, No. 23 (December 9). Washington, D.C.: US Department of Commerce.
_____. 1941b. Japanese Population in Selected Cities of the United States. Sixteenth Census of The United States. Series P-3, No. 24 (December 10). Washington, D.C.: US Department of Commerce.
_____. 1941c. Japanese Population for Each County in the Pacific Coast States by Sex and Nativity or Citizenship. Sixteenth Census of The United States. Series P-3, No. 25 (December 11). Washington, D.C.: US Department of Commerce.
_____. 1941d. Population: Japanese Population in Selected Counties and Cities of the United States by Sex and Nativity or Citizenship. Sixteenth Census of The United States. Series P-9, No. 5 (December 19). Washington, D.C.: US Department of Commerce.
_____. 1942.a. "Bureau of the Census" In Report of the Secretary of Commerce, "1942 Report by the Bureaus." Washington, D.C.:US Government Printing Office.
. 1942b. "Exhibit A - Work done on reimbursable basis by Bureau of the Census for war or other agencies charged with execution of some phase of the war program. December 1941 through November 1942." December 4, 1942. Mimeographed. War Work Reports folder; National Defense; Papers of Philip Hauser (item 146); US Census Bureau, Record Group 29; National Archives Building, Washington, DC.
_____. 1943a. Uncle Sam: How He Grew. 150 Years of American History As Recorded in the U.S. Census. Washington, D.C.: US Government Printing Office.
_____ . 1943b. Sixteenth Census of the United States: 1940, Population, Vol. 2, Characteristics of the Population (Washington, D.C.: Government Printing Office.
. 1943c. "Meeting of the Special Advisory Committee to the Division of Vital Statistics, Bureau of the Census, January 8, 1943, Agenda and Minutes." Mimeographed. Exhibits used with the Advisory Committee Meeting, January 22-23 1943; General Records of the Census Advisory Committee (item 148); US Census Bureau, Record Group 29; National Archives Building, Washington, DC.
_____ . 1947. Fact Finder for the Nation. September 1947. Washington, D.C.: US Government Printing Office.
_____. 1974. Bureau of the Census Catalog of Publications, 1790-1972. Washington, D.C.: US Government Printing Office.
_____. 1979. Twenty Censuses: Population and Housing Questions, 1790-1980. Washington, D.C.: US Government Printing Office.
_____. 1980. Census '80: Continuing the Factfinder Tradition. (By Charles P. Kaplan, Thomas Van Valey, and associates) Washington, DC: US Government Printing Office.
_____. 1982. "Statement on Census Bureau Actions at the Outset of World War II as Reported in Infamy: Pearl Harbor and its Aftermath by John Toland." October 1982. [Attachment to Chapman, 1982.] CWRIC microfilm.
US Commission on Wartime Relocation and Internment of Civilians (CWRIC). 1982. Personal Justice Denied: Report of the Commission on Wartime Relocation and Internment of Civilians. Washington, DC: US Government Printing Office.
_____. 1984. "Papers of the US Commission on Wartime Relocation and Internment of Civilians." Microfilm of documents in National Archives Record Group 220.18.25. 35 reels. [Sometimes referred to as CWRIC microfilm]
. 1997. Personal Justice Denied: Report of the Commission on Wartime Relocation and Internment of Civilians. Seattle: University of Washington Press. Reissue of CWRIC [1982], with a new forward by Tetsuden Kashima.
US War Relocation Authority. 1945. "Administrative Manual - Index I." Release # 190, dated 5/5/45 and covering memorandum from the Assistant Director. [CWRIC microfilm]
. 1946a. Administrative Highlights of the WRA Program. Washington, DC: US Government Printing Office.
. 1946b. The Evacuated People: A Quantitative Description. Washington, DC: US Government Printing Office.
. 1946c. Impounded People: Japanese Americans in Relocation Centers. Washington, DC: US Government Printing Office.
. 1946d. Wartime Exile: The Exclusion of Japanese Americans From the West Coast. Washington, DC: US Government Printing Office.
Van Arsdol, Maurice D. Jr. and Aubrey Wendling. 1995. "Calvin F. Schmid." In Footnotes 23 (January): 13.
Weglyn, Michi Nishiura. 1996 . Years of Infamy: The Untold Story of America's Concentration Camps. Updated edition. Seattle: University of Washington Press. [Original edition published in 1976].
Welpton, P. K. 1938. Needed Population Research. Prepared under the auspices of the Population Association of America. Lancaster (PA): The Science Press.
Young, Allyn W. 1918. "National statistics in war and peace." Quarterly Publications of the American Statistical Association 16 (121): 873-885.